Roche Holding AG (ROG) Earnings Call Transcript & Summary

Ladies and gentlemen, welcome to Roche Virtual Pharma Day 2021. My name is Henrik, and I'm the technical operator for today's call. Kindly note that the webinar is being recorded. [Operator Instructions] At this time, it's my pleasure to introduce you to Karl Mahler, Head of Investor Relations and Group Planning. Karl, the stage is yours.

Okay. Then let's get started. Welcome to our traditional Pharma Day. Warm welcome from my side. I'm just checking the attendance list. So it's still piling up, but we do expect today more than 700 people. So that is really a considerable high interest in the day. So thanks for your interest in Roche. You will have a possibility to ask questions via the phone, via the Zoom chat, also with the e-mail, so you can also drop an e-mail on the [email protected]. And I trust that you had an opportunity to download the presentation by now. Please note that there are also some API slides in the same link, respectively, on the website. I know that many of you use these API slides not only during for this event, but throughout the year so this is in high demand. And last but not least, we also do have a survey. So we want to be also a learning organization here. So in the second part of the day, there will be a pop-up of 10 questions. And if you don't like to answer the questions, then just move it to the site. This is perfectly fine. If not, then we would be extremely pleased if you could kindly help us improving ourselves and the event for next time. So this is the program for today. So we are basically 2 parts. First part, Bill, will open with a strategy. He will give us reasons why we aim to invest more in R&D, while we believe this is the right thing to do at the moment and why we believe it's the right thing to reshift the more resources into R&D. Then we have Teresa. So basically, she's talking about current funding, where the current funding of the strategy is coming from. She talks about launch excellence, product differentiation, with a focus clearly here on main drugs ongoing. And we will have a short break of 5 minutes, just a kind of a mental break. Then it goes into the second part, Levi and Paulo will give us an answer on the questions where do -- where we see our pipeline products making a difference to patients. Where do we see the standard of care moving? Why we invest at the moment in which trial, in which asset? And how we can contribute to the changing standard of care? And then we have 2 call-outs, one will be on ophthalmology by Nilesh and one from Barry on the infectious disease, and interestingly infectious diseases, many companies have said goodbye. They are not invested in. We have still a commitment here as we had for CNS many, many years ago, and you know that we have a leading pipeline at the moment in CNS by just focusing on science and on the long term and on the right things. In terms of day, we try to be really 50-50 this time, 50% presentation time, 50% we will have Q&A. And before we go into the event, maybe 2 slides from my side. This is the first 1 on this. It shows a bit history of Roche 2011 to 2021. The past 10 years, you can see basically 3 things on that slide. First of all, pharma was growing by 5 billion, which is a lot. Second is, you can see that we had really a focus on oncology, as 55% to 60% was invested in oncology 2011. It's now less, maybe 40%. AH&R; Avastin, Herceptin, Rituxan were about 55% to 60%, maybe 2011. It's now down to 20. For the group, it's maybe 10, 15. So you can see the huge reshift in what we really -- what the pharma division could achieve during this time by reducing the risk. Second is that there is a clear product portfolio rejuvenation and diversification ongoing, and you can see it here in the colors by having less orange and much more for immunology, neuroscience and other therapeutic areas. And why this was the case and why we could achieve this one? You can see on the left side, there was a continuous focus on science. We are the company with 38 breakthrough therapy designations, which we received since 2013. 8 Immunology, 3 in Neuroscience, Oncology 25 and Ophthalmology 2. So there's also a good rejuvenation and diversification in the innovation part testimony. And on the right side, it's kind of an external -- yes, call out here, testimony from XXXXXXXXXXXXXX], which was published last week, which shows also here the refreshed outlook for the Roche Group for pharmaceutical Roche. And with this one, I would like to hand over to Bill. Bill, over to you.

Thanks, Karl, and thanks all of you for joining us. I think we've been really excited to have an opportunity to talk about the progress that we're making. And to explain a little bit about what we're doing? Why we're doing it? And so let me just start, you see the title of my presentation, sustainable growth by delivering more patient benefits at less cost to society. And many of you know that about 2.5 years ago, in early 2019, we committed to really a fundamentally new strategy for Roche Pharma, wherein we said -- we started by looking at what does society need? And it was really clear that there's still a lot of diseases with high unmet needs. And patients need better medicines. They need targeted medicines. They need medicines that are curative. They need medicines that really solve their problems rather than giving them 6 weeks or 2 months of additional life that cancer patients would have a cure, that babies born with spinal muscular atrophy would have new options and ability to live a normal life. And the people with hemophilia would have also really a new day to move beyond the tired old medicines that they've had to rely on for so many years. And we said -- so how do we do this? And we said, we fundamentally have to do 2 things. We have to invest a lot more in R&D. And in order to do that in a reasonable way, we have to spend a lot less kind of on everything else. We also need to be much better, more efficient, more effective in R&D. And that has to do with personal productivity. It has to do with technologies. It has to do with really moving beyond the 20th century means of developing medicines into things that take full use of the evolution of technology. And then the other thing we need to do is we really need to create a workplace where each person at Roche Pharma has the opportunity to make an impact every day. And in large organizations too often, people are kind of bogged down in the bureaucracy, in the approvals and going through management levels and that -- and we said, we're not going to play at changing that anymore. We're going to make radical changes and we began that journey, and it's already begun to pay off. And I hope you'll see that as we present today. But let me move on and show you some of that progress. So first off, let's talk about what's changed in the last year since our last Pharma Day presentation. First off, at the high level on the late-stage pipeline, really pleased at the evolution here. We now are up to 18 medicines in Phase III and registration, which is a new all-time high. You can see the recent past. This is almost an 80% increase over what we've been doing in the last decade. And I think you'll also agree that a lot of these medicines -- they're not average medicines. They have tremendous potential. And in the earnings results that we gave in July for the first half, we announced that we had begun 12 new trials. We've got an additional 10 new Phase III studies in the second half of the year. And so basically, we're talking more than 20 new Phase III studies in a single year. So not only do we have more medicines in Phase III in registration, but we're beginning more pivotal studies than ever before. And again, if you look at the list, it's quite impressive the things that we're taking on. Now we've also talked for a number of years about filling the gap. The gap created by the loss of IP on Avastin, Herceptin and Rituxan. And so the last couple of years, we've given this update. And basically, this is looking at the analyst consensus and what's the difference between our projected launches and our projected losses. And you can see in 2018, we had about 7 billion more to gain than we had to lose. Last year, that number had grown to 8 billion, and now it's basically 12 billion. So we're really pleased at how we've made progress despite the decline in sales of AH&R, really tremendous growth of the pipeline of marketed products, and we really, I think, are moving past the point where biosimilars is a major factor for us as we look forward. Now transformation, I mentioned, how do we help all of our people have a chance to do something important for patients every day. And this is really a major commitment for all the people of Roche. And the leaders of Roche are really strong behind this. It's been an ongoing effort. I mean this slide highlights the areas where we're focused, which is basically everywhere across pharma. It comprises several parts. There's a new way of leading instead of command, control that we call VACC. It's vision, architecture, catalyst and coach. It's a very different way of operating. It's required all of us and me included to basically relearn what leadership looks like. People come to work at Roche to do something great for patients. They don't come to work to be managed. And there is a need for vision and for clarity and for structure that helps people get things done. But people have come here to be babysit. And this leadership approach we find is really a whole new day in terms of empowering teams in a truly dynamic way that allows them to get on and make the best choices. Examples of things that, that's achieved, for example, in regulatory, it used to take us 26 weeks to file. I mean we have the most filings of any company in life sciences, regulatory filings. It used to take us 26 weeks on average. Now we're down to 13 weeks. And we tried that for 10 years to improve, and we couldn't until we really turned people loose and give the teams the power to make decisions. Other parts, operating principles, new ways of allocating resources, moving beyond the tired old budget formulas and then outcomes-based planning, which is a powerful way for people, teams to align their objectives and to drive things forward faster than ever before. And the whole goal of this is to deliver our vision of many more patient benefits at a lower cost to society but also to make for better jobs. Better jobs is happier people. It also means that we have -- we maintain our industry low attrition. And I think it's basically a win for the mission. It's a win for patients, and it's a win for our company. Now if that all sounds inspiring, but like where's the proof? Let me share. And by the way, this chart is basically lined up in the order of when different groups began their transformation work. So starting on the left with Pharma Technical, that's our manufacturing and quality organizations and process development and Pharma U.S. These were groups that started transforming in 2016 already. So it's been 5 years. And what you can see is, for example, 68% volume growth with 18% lower headcount today than we had when that began. And now fortunately, we've been able to do a lot of that through just sort of normal attrition, but the net result is basically a doubling of product output per person. So that means something. I think that's a pretty clear sign that the transformation is yielding results. In pharma in the U.S., 23% sales growth, head count is 19% lower. If you do the math on that, that's a -- it's more than a 50% increase in sales per commercial employee. In Pharma International, 18% sales growth, 15% lower headcount. And again, this is close to 40% increase in per person productivity. And now you see in China and Pharma Development, a little newer on the curve. But again, impressive figures. In Pharma Development, we have a 44% increase in new molecular entities at pivotal stage and we're doing that with 27% more headcount. And I think you'll hear from Levi about our efforts to improve on that. Now what do we do with the benefits of empowering people, of allowing people to accomplish more every day, we can basically invest that in more projects and more R&D efforts from -- beginning from research all the way through. So here, you can see our P&L for the first half. And what you see here is a reduction in cost of sales of 3%, reduction of marketing and distribution of 6%, G&A down 2%, R&D up 19%. By the way, this isn't -- this pattern won't endure through the year quite that starkly. We had a large increase in R&D, partly driven by the COVID therapeutics effort where we have 3 therapies that we've been studying for COVID in Phase III -- in Phase II and Phase III studies. But on the right, you see the net effect. We're talking about fundamentally changing the nature of the company to be much more lean and efficient and customer-centric rather than product-centric on the commercial side, but to deliver a lot more investment in R&D. So let's talk about what that's buying us? What do we get for that investment? And so first, in terms of the strategy, we look at the health care field, and we still see many, many diseases with high unmet need. Patients need better therapies. They need less toxic therapies, more convenient therapies and mostly they need cures. And if you look, areas like oncology, CNS, cardiovascular, anti-infectives, I mean, still very large unmet needs, but also many, many other therapy areas with big needs. We're approaching this with many modalities. So on the left, you can see the industry pipeline by therapeutic modality. And what you'll notice at the industry level, the modalities have been shifting. Back in 2000, it was 60% small molecules. And now industry-wide, it's only 50% small molecules. If you look at Roche and our pipeline, our pipeline and our modalities have actually evolved much faster than the industry. And you can see there, it's quite a balanced picture with a lot of different modalities. And basically, our approach is we're in every modality when we believe it's the best modality for a serious disease. And so sometimes we get asked questions like, oh, what about this modality? Why aren't you there? And the answer usually is, well, because we have something else that we think is better in -- for that disease. But when we see the opportunity, we tend to move. Here's -- again, for your reference, a listing of some of our key modalities and the diseases we're targeting with them, I won't belabor it. But I think, again, it's -- it reinforces the point that Roche is in every modality. And if we're not in a modality today, if it proves important, you will see us in it tomorrow. We're also, I would say, very agile with respect to deals. We do a lot of early-stage deals. We don't believe in kind of waiting to Phase III and then rolling the dice with big transactions. We don't think that's worked very well for large companies in the past. It's definitely a seller's market at that point. And we believe in establishing early partnerships where we're really able to bring the full resources of Roche scientific development, manufacturing to bear on the program and to get in early enough to do that. I think you'll also see that we're very disciplined. So for example, in the 12 months ending last year, we had done 5 late-stage deals, okay? Since that time, we've done 0 late-stage deals. And why is that? Because we found 5 late-stage deals that were both strategically important, great science that we liked, but also they made sense financially. In this year -- in this calendar year so far in 2021, we haven't seen 1 that met all those criteria, and we don't feel pressure to push the button on deals that frankly don't make financial sense, even if they look good strategically and scientifically. So let's talk a little bit about the late-stage portfolio and the evolution. You're going to hear a lot more during the day today. But high-level perspective, in 2018, we had 10 blockbusters as a company. The consensus estimate for 2021 is that we'll have 16 blockbusters this year. And now we have an additional 24 molecules that are either already approved or that are waiting in the wings in Phase III, will be approved in the next few years that have large sales potential. So I think this is one of the reasons we're so excited about what we can bring to patients around the world at Roche in the years ahead. A few examples. Ophthalmology, an area that we've been really pioneers of bringing biologics into ophthalmic indications beginning with Lucentis back in 2006. Now we've got basically 2 late-stage programs. They're in filing mode right now. Faricimab and the port delivery system with ranibizumab. Both of these are going to be approved before this meeting next year, and they have a real opportunity to provide a big benefit for patients and to be important medicines for Roche. And I think it's also important to add that right now, Roche's footprint in Ophthalmology is limited to the United States; whereas with the programs, we will have these medicines around the world. So it's a great day for us to move forward. But beyond that, we also are working on new molecules for the port delivery system, the so-called DutaFab program that you'll hear more about today. And we're also working on a number of collaborations early on geographic atrophy, which is a really high unmet need area still in Ophthalmology. Central nervous system diseases, again, very high toll on society, on people. And we're really coming at this, I think, with a multidisciplinary approach. It's not only about the science or the medicine, but also, for example, digital tools, imaging, patient-enabled tools, cell phone-based tools. And so you can see this chart gives you an example in MS, in SMA, in muscular dystrophy, a number of areas where we're advancing not only the science, not only molecules but also tools that can help physicians diagnose patients better, help indicate which therapy is suitable or when a change in therapy is necessary, but also help patients to really monitor their disease and to know what benefit they're getting from their current medicine or a new medicine. We're also going early. I mentioned that our goal was to deliver 3 to 5x -- our internal goal is to deliver 3 to 5x more patient benefit at half the cost of society. Part of how we achieve such an ambitious thing is we have to deliver more cures. It's not good enough to extend life by 2 months. We need people to go to that last appointment with the oncologist, not the last appointment because they face death, but the last appointment because they face life without cancer -- life beyond cancer. And so we're investing very heavily in this. You can see these examples. So for example, in lung and rare tumor types, these are all studies, either an adjuvant, neoadjuvant, stage 3 opportunities for cure for a sizable portion of the population. And we're really pleased to have already demonstrated that Tecentriq works in early-stage lung cancer and that we can achieve a long time of progression-free survival there or disease-free survival. Breast cancer, we have a number of molecules including Tecentriq in TNBC and triple-negative breast cancer, but also giredestrant that we're looking at a broad range of settings, but this is in estrogen receptor positive cancer and an area that's 70% of breast cancer. But also, again, in GI and gastrouinitary (sic) [ genitourinary ] cancers and in hematology with POLARIX with the readout recently announced. So the last thing I want to mention is what we're doing to make a sustainable impact. I hope you'll agree that the very nature of our vision speaks to sustainability, delivering more benefit to patients at less cost to society. And frankly that's the biggest thing we can do to be a sustainable health care company, is to deliver what people look to us uniquely to deliver. But we're not content with only that. And so we're looking at a number of other metrics. We're very pleased to be the most -- to be ranked as the most sustainable health care company by the Dow Jones Index for the 11th time. We work hard at that. We've worked hard at that for over a decade. And we will continue to earn our place. For example, on CO2 emissions per employee, you can see the chart on the left there, where we brought them down by about half over the last 8 years. And that came through concerted investments, clear strategies and really a will to change. And I want to make clear, this is not by buying credits, okay? Buying credits is I think that's debatable what good you bring to the planet. I think reducing emissions, there's no debate, and this is pure reduction of emissions. Also in terms of access to our products. And we have international differential pricing around the world so that patients in countries with lower income are able to receive our medicines at a much lower price. In the U.S., we've been very careful about price increases. We basically had our net price increases in line or below medical inflation in the U.S. for a decade. You can see the most recent years here. And then in terms of things like diversity and inclusion, we think that's really important, not only as a contribution to society, but it's just a fundamental right for all of our employees to feel like this is their company. It's a place that belongs to them and they belong here. And so you can see, as an example, our progress in women, in workforce and women in leadership. We've increased the number of women in senior leadership by over 50% in the last 6 years. I think that's good progress. And we think there's tremendous gains ahead on that, both in the name of gender parity, but even more than that in the name of performance. So I hope you're convinced that Roche -- that we mean business when it comes to sustainability. Now what can you expect for us going forward? A couple of things. We're going to continue to enter new franchises, new areas of medicine, new areas of unmet need. I think our plan that we set out 10 years ago when we faced the looming introduction of biosimilars, we came up with a plan, how we were going to handle that. You can sort of see that in the blue bars in the middle. But then if you look at sort of the next bar over to the right, the peach color, these are sort of additional areas that we've entered beyond the areas affected by biosimilars, and there's just more ahead. And if you look at the readouts that we have in 2021 still to come, 4 more significant readouts and then many, many readouts in 2022. I think we've listed 13 that we think are particularly notable. That's basically every 4 weeks on average in 2022. Three pivotal readouts for tiragolumab, giredestrant our first pivotal readout for our estrogen receptor degrader program and then the port delivery system in DME, gantenerumab in Alzheimer's and then a number of important readouts in early cancer for Tecentriq and Alecensa. So again, a big year ahead. We're really excited to have that opportunity to bring the pipeline forward. And then lastly, I just want to say, we've reiterated our positive outlook. We said a few years ago, we believe we can grow through the biosimilar impact time, and we still believe that. All indications are go, the pipeline continues. The progress of our launch products remain strong, and we look forward to continuing to deliver great news for patients, but also for investors in the years ahead. So with that, I want to hand it over to our Head of Global Product Strategy and our Chief Marketing Officer, Teresa Graham.

Great. Thank you, Bill. Good afternoon, everyone. Normally, in my presentation, we would take a pretty comprehensive look through the full late-stage portfolio. But given the sheer amount of activity that's happening over the next 12 to 18 months, we're really going to focus in on the near growth drivers, the things that as Bill mentioned, are going to help us fill that gap. Before we jump in, though, I did want a moment and just reflect on what COVID-19 impact has meant to health care systems around the world. We've seen variable continued impact across geographies and across different therapeutic areas over time. In Oncology, for example, patient visits have by and large returned to pre-pandemic levels. We are still seeing a slight depression in the Hematology patient visits. And contrast that with multiple sclerosis in the U.S., where those switch numbers are still 10% to 15% below what they were in pre-pandemic times. With the emergence of the Delta variant in certain parts of the world, we are seeing sort of a renewed impact, particularly in our ability to see physicians. But we are still really being able to focus on all of our launches and delivering our great medicines to patients. As Bill mentioned, we've had some really significant news flow. Just over the last year, the adjuvant non-small cell data for Tecentriq, POLARIX, for Polivy, the AMD data for faricimab and then a series of trials that have reinforced the benefit of Ronapreve in COVID-19 patients. And coming up shortly, we'll have the first pivotal data for our 2 bispecifics, mosun and glofit. Bill touched on many of the exciting data sets that we have reading out in 2022, including all of the adjuvant and neoadjuvant data for Tecentriq. We also have an adjuvant ALK positive non-small cell lung trial cancer ALINA reading out for Alecensa. The first pivotal data for tiragolumab and giredestrant, and then the very hotly anticipated gantenerumab data in Alzheimer's. So now let's take a more deep dive into some of the individual opportunities, and let's start with Tecentriq. Tecentriq already has annualized sales of more than 3 billion with some significant near-term catalysts. Near-term growth both in '21 and '22 will be largely centered around driving the launch of adjuvant non-small cell, particularly in the U.S., as we continue to expand in ex-U.S. and geographies that have not yet launched for small cell and TNBC. Next year, we have the 3 readouts for 4 trials in early disease, including head and neck, RCC, HCC and then the neoadjuvant study in non-small cell Additionally, there still remains high unmet need across tumors where PD-1 and PD-L1 have been approved and first-line non-small cell, for example, we still see that overall survival is less than 2 years. And that provides a great opportunity for next-generation CIT combos like Tecentriq and tiragolumab to really reset the standard of care in markets where Tecentriq may not already be the market leader. So let's talk a little bit about adjuvant non-small cell. So we presented very exciting IMpower010 data for Tecentriq in adjuvant, which was followed under the real-time oncology review with the FDA as well as Project Orbis. The granting of RTOR really speaks to be unmet need in this patient population as well as the very compelling data. I think many of you already know that adjuvant is still an evolving treatment landscape and currently only about 30% of non-small cell patients are diagnosed with resectable disease, and treatment rates continue to be modest, primarily because of the limited benefit and toxicity of chemotherapy for these patients. We do expect as more treatments come online like Tecentriq, like potentially EGFR -- for EGFR-positive patients, for ALK-positive patients that we will see screening increase, we'll see testing increase and ultimately, we'll see systemic therapy increase for this patient population. Moving on to tiragolumab. I think Bill mentioned this as well. We have 9 Phase II or Phase III trials initiated. We have the broadest and most advanced clinical program in tira with 4 readouts planned for next year. These readouts allow us to not only build on places where Tecentriq is also strong, but to compete in new indications with head-to-head trials against standards of care, including durva and stage III, non-small cell and pembro plus chemo and first-line non-small cell. The market opportunity for tiragolumab is significant, particularly when we look at the lung cancer where PD-L1 and PD-1 to date accounts for $10 billion in annualized sales, and we believe that this is where a combination such as Tecentriq and tira has the opportunity to reset the standard of care. Switching over to breast cancer. We continue to innovate for patients with HER2-positive breast. We continue to see solid growth for Kadcyla in the adjuvant setting, where we've established an extremely high bar, both in terms of efficacy and safety. We're also seeing really encouraging uptake for Phesgo, both in the U.S., particularly in academic settings, but also in other parts of the world that are typically more receptive to subcu formulations. For example, in the U.K. right now, we see 1 out of every 2 patients actually receiving Phesgo, which is really encouraging. We are also continuing to build on the standard of care. So the graph that you see here is the PD-L1 subset of the KATE-2 study, where we saw very encouraging signs for both PFS and OS, which has encouraged us to actually launch some combinations with Tecentriq both in combination with Kadcyla and Perjeta both in metastatic and early breast settings. Moving on to giredestrant. As you know, and I think Bill mentioned this as well, hormone receptor positive breast cancer is an extremely large opportunity comprising about 65% of all breast cancer patients. That's 3x the population of the HER2 market, in hormone receptor positive breast cancer endocrine therapy including fulvestrant aromatase inhibitors continues to be the standard of care across both early and late line settings. And it is our hope that giredestrant will replace endocrine therapy across all lines in the future. There remains significant unmet need in this patient population, up to 50% of early breast cancer patients will actually stop their treatment early due to toxicity. And about 30% of patients will ultimately develop metastatic disease. We do believe that giredestrant has the potential to be best-in-class, both because of its differentiated MOA. Because of the high potency, we believe that we're 7 to 15x more potent than other SERDs in development. There was just recently a paper published in the Journal of Medicinal Chemistry that Levi will talk about in his session. We're well tolerated, both in mono and combination therapy, and we are able to have a standardized dose, both for mono and combo because we don't see those drug-to-drug interactions. We have the broadest clinical program out there with SERD, including a head-to-head trial with standard of care in adjuvant. So lots of really exciting data coming for giredestrant, and we do believe this has the opportunity to be a game changer. Speaking of game changers, Polivy plus R-CHOP, the first positive trial in first-line DLBCL in over 20 years since the introduction of R-CHOP. I think as everyone knows, first-line DLBCL can be a curative setting. But unfortunately, about 40% patients are still not cured with R-CHOP, in the first line. And that means that patients that either are resistant or relapsed unfortunately have a relatively poor prognosis. This is precisely where we believe that Polivy can play. This is a very large multibillion dollar market opportunity. There have been no new first-line therapies approved since R-CHOP. And we do believe that we have the opportunity to really change the standard of care. We're excited to present the Polivy data later in the year, and we are actively discussing with health authorities about how to bring Polivy to patients as fast as possible. Also in the Hematology space, we've got mosun and glofit, our CD20 and CD3 bispecifics, both of which have the potential to be first-in-class and best-in-class in follicular lymphoma and DLBCL. Mosun continues to demonstrate high and durable response rates with a very manageable safety profile, no hospitalization, which makes it very attractive across a broad range of indications and treatment lines, so particularly things like follicular lymphoma or elderly unfit patients. For glofit, it has the opportunity to be the best-in-class efficacy comparable to what we see with the CAR-Ts. We are seeing in our latest data that step-up dosing can actually help to manage higher target doses while maintaining a manageable safety profile. So really excited to see the first pivotal data out of both of these drugs a little bit later this year, the most in pivotal cohort in third line plus follicular, will file at the end of this year, and we expect the glofit pivotal cohort in third-line DLBCL to file at the -- in early 2022. We have also launched randomized Phase III trials in both relapsing and reletting follicular lymphoma for mosun and in second line plus DLBCL with glofit in combinations with standard of care. And we are actively thinking about how to move into first-line DLBCL in combination with Polivy. So moving on to nonmalignant heme, Hemlibra continues to be transformational for hemophilia A patients regardless of inhibitor status and age. We see the 3-year nobly treatment rates for Hemlibra being in the 80% to 90% range compared to 40% to 50% for factor VIII. We continue to gain market share both in the U.S. and globally. Our non-inhibitor approvals are now at 90 -- more than 90 countries with reimbursement in more than 30 of those countries. And just for a little bit of context in the inhibitor space, where we already have very strong penetration for patients, we see reimbursement in 90 countries. So we would expect that we'll be bringing a lot more countries online in the very near future. We also have a significant amount of data that will be coming out for Hemlibra next year, starting with HAVEN, 6 which is the mild to moderate study that we ran in the EU. We hope that this will not only inform a label update for mild to moderate patients in the EU, but will continue to provide reassuring data for those patients who are being treated around the world. We have the 5-year follow-up data from HAVEN 1 through 4, which are also expected next year. And then very importantly, we have the HAVEN 7 data in which Hemlibra will be the first novel therapy to be studied in the prophylaxis setting for infants also expected next year. So just Hemlibra continues to be a really important and foundational drug for the treatment of hemophilia. Now I won't spend too much time on Ophthalmology because you have Nilesh coming up in a little bit. But in any conversation about how we are approaching commercial opportunities for the next 18 months, it would be really remiss not to talk about ophthalmology. We're preparing for our first launch in PDS in 2021, in faricimab at the beginning of next year. And I think as everyone knows, this is a very significantly large and growing market expected to reach up to $15 billion in 2025. But we still see a significant amount of unmet need despite the availability of treatments largely because patients just can't get the level of injections in the real world that they need to maintain optimal vision outcomes. And this is where both faricimab and port delivery system really come in. Faricimab is the first new MOA in AMD and DME in more than 15 years and has shown strong durability with the majority of patients being -- about half of patients being able to be maintained on Q16 dosing. For those patients who are already responsive to anti-VEGF therapy, but for whom can't be maintained on a current extended treatment regimen or just simply can't manage the logistics of going into the office as frequently as they need to, the port delivery system is a wonderful alternative to maintain steady vision while with almost all patients being able to be maintained on dosing every 6 months. As Bill mentioned, we now have the global rights for both faricimab and for PDS. And we're really looking forward to these launches. These are physicians that like to switch. They like to try new things in their armamentarium. We would expect a reasonably fast adoption for faricimab, and it's as I think, again, as Bill mentioned, with port delivery, we've got a go slow to go fast strategy here to make sure that we're managing the introduction of a new kind of surgery. Moving on to OCREVUS. OCREVUS continues to have a very strong growth profile. In the U.S., we are seeing the anti-CD20 class grow to now over 50% of patient, but with significantly more room to expand. Within the anti-CD20 class, OCREVUS continues to have the best in disease efficacy and safety. We remain the only medicine with a consistent reduction of disability progression across all endpoints and trials. We're the only therapy approved in PPMS, and the higher-dose OCREVUS studies look to further improve on that best-in-disease profile. The twice-yearly dosing drives better compliance. We see a greater than 90% persistence of compliance after a year, which is superior to what we see whether they're oral and injectable medicines and shorter infusion is continuing to gain traction in the markets in which it's approved. Moving on to one of the drugs that we launched last year Evrysdi. Evrysdi is the first and only oral SMA treatment with meaningful efficacy in the broadest SMA population ever studied in pivotal trials. We see patients in all SMA types in a broad range of ages, both treatment-naive and previously treated coming on to Evrysdi, and we have had the fastest uptake for DMT with our launch in the U.S. That strong performance in the U.S. is being mirrored with what we're seeing in our ex U.S. launches with about 20% market share in Germany just within 4 months of launch. We have ongoing dialogues currently happening with regulatory authorities throughout the EU, and we expect to be able to bring Evrysdi to many, many patients in the future. The global SMA market continues to grow, and we expect it to be greater than $5 billion by 2025. And in many cases, across the globe, there are significant numbers of patients, who are significant percentage of SMA patients that are currently untreated. The profile of Evrysdi makes it attractive both for patients, who are not yet on treatment as well as patients who may like to switch their therapies. And so we see a lot of really good things ahead for Evrysdi. And last but not least, gantenerumab. So with gantenerumab, we will bring the most comprehensive data set in Alzheimer's with the GRADUATE studies, which we'll readout next year. We're very confident that with GRADUATE, we will be able to deliver a clear and robust data set. We have two well-powered parallel studies with about 1,000 participants each. We have extended trial duration, the longest duration out there of 27 months. We have optimized titration in the single dose, and we have demonstrated plaque reduction, particularly in our OLE where 80% of patients were below amyloid positivity at 3 years. Also importantly, we will be the first and only subcutaneous treatment for Alzheimer's disease. Subcu allows a treatment flexibility, which is not only just important to patients, but it's important to health care systems as well, reducing the IV burden of infusions for patients and for health care systems. As we would do with any of our medicines, we are in conversations with health care -- with regulators around the world to learn how we can bring this drug to patients in a way that makes the most sense. And so with that, I think I am actually turning it over to Karl for a break.

Yes. Thanks a lot to both of you. We will have a short as announced 5 minutes break -- a mental break before we go into the detailed analysis of trial data and so on. So we reconvene here at -- in 5 minutes, so 50 -- 2:50 over time CET. [Break]

Thanks a lot. So we're coming to the second part of the day. Actually, we have 3 locations today. One is in Basel. One is in South San Francisco. So well -- it's an awful time for you. Levi, I guess, you're used to know since you have overseas 9 hours difference. But thanks that you put the burden on you to wake up so early. And then we have Barry sitting in Berlin. And I'm always wondering how well that works with all these technical interlinks and so on Levi, over to you. Thank you.

Yes. Thank you, Karl, and hi, everyone. It's always a pleasure to discuss our pipeline. And actually, Karl, we actually have 4 locations because I'm pleased to recognize our new Head of Global Oncology and Hematology Development, Charlie Fuchs. He is also with us virtually today from -- still from Boston, Massachusetts area. So you'll have a chance to hear from Charlie during the Q&A session. Next slide, please. So our oncology pipeline remains among the strongest in the industry. And since the Pharma Day a year ago, our portfolio has continued to deliver the kinds of medical advances and patient benefits that Bill talked about. So I'll provide several examples that illustrate some of those strengths, starting with our heme malignancy portfolio, where we have exciting data for several assets. And then on the solid tumor side, I'll touch on recent developments in our breast cancer program, and then I'll review the progress of our adjuvant programs as well as some of our new efforts -- and newer efforts, I should say, in immuno-oncology. And then finally, I'll conclude with updates from our benign hematology programs. Next slide. So this is a high-level view of time lines for pivotal data readouts from our late-stage pipeline. And as you can see, already this year, we've had 3 of 5 key pivotal readouts and they've had positive results 2 of these, the Tecentriq adjuvant lung result and Polivy in first-line DLBCL represent significant new opportunities for these medicines. And in 2022, we anticipate at least 8 key pivotal readouts. And 5 of these will be new molecular entities, and those include glofitamab, which is 1 of our CD20-CD3 bispecific antibodies. And by the way, another bispecific, mosunetuzumab we'll put out later this year. And then we also have tiragolumab, our anti-TIGIT immunotherapy asset. We have giredestrant, which is our oral SERD. We'll have overall survival data for ipatasertib, which is our AKT inhibitor, together with abiraterone, a metastatic caster resistant prostate cancer. And you'll recall that study was already positive for PFS in patients, who were PTN-negative by IHC. And then we will have crovalimab in paroxysmal nocturnal hemoglobinuria from a -- 1 of our Phase II trials to readout, the 1 being run in China. And then we also have additional indications for Tecentriq and an adjuvant study of Alecensa reading out. So it will be a really interesting year for our pipeline next year. And then, as Bill mentioned earlier, when we look out beyond 2022, we have many more key pivotal readouts on the horizon. So we think our oncology portfolio should remain industry-leading for years to come. Next slide. And when we say industry-leading, we mean both quantity and the diversity of our pipeline innovation. So in oncology, for example, we have long since moved beyond small molecules. I mean, Bill illustrated this earlier, but also even monoclonal antibodies and ADCs, where our portfolio now includes bispecific antibodies, protein engineering innovation and cutting-edge platforms such as neoantigen vaccines and personalized T cells. But through it all, the mantra is always to follow the science. So we allow disease biology insights to guide development of first-in-class and best-in-class treatments that improve outcomes. Next slide. And I would say nowhere are these principles more evident than in our hematologic malignancy pipeline. And for example, in both indolent and aggressive B-cell malignancies, we brought forth breakthrough medicines on multiple occasions ever since rituximab in the late 1990s. And most recently, this has been evidenced by Venclexta in CLL, now Polivy in first-line DLBCL and we have reason to believe that our portfolio will continue to change the standard of care in B-cell malignancies, in particular, our bispecific antibodies are showing real promise here. Next slide. And furthermore, we are expanding into other malignancies such as acute myelogenous leukemia, multiple myeloma, myelodysplastic syndrome. Specifically, we're developing Venclexta in additional AML context and a first-in-class bispecific antibody called cevostamab in multiple myeloma. So we'll talk a little bit more about that shortly. But overall, these efforts should allow us to benefit patients in disease areas where we've not previously had a presence and thereby extend our heme malignancy impact. Next slide. All right, Let's take a closer look at a couple of these opportunities, starting with Polivy. So in the POLARIX study, which has already been mentioned by Bill and Teresa and of course, this readout positive recently. We tested the hypothesis that giving Polivy together with rituximab and CHP, so in essence, swapping out Vincristine for Polivy in the R-CHOP regimen, we tested whether that might prove superior to R-CHOP in the frontline DLBCL treatment setting. And despite the fact that more than 20% of DLBCL patients eventually relapse the refractory to treatment, R-CHOP has actually stood as the standard of care for more than 2 decades. And there were at least 11 prior attempts to improve on R-CHOP, all of which were unsuccessful prior to this Polivy-based regimen. So R-CHOP had already set a high bar in DLBCL treatment, and it stood for a long time. But this -- what we call Pola R-CHP regimen has shown an efficacy and safety profile that could potentially bring a new standard of care treatment for newly diagnosed DLBCL patients, and so we'll be presenting these results in an upcoming meeting, but we're obviously very pleased with that outcome. And meanwhile, we continue to pursue a robust development program with Polivy and that will include a Phase III trial, the SUNMO study, in combination with mosunetuzumab in second-line or greater DLBCL and then multiple other combinations with agents from our heme portfolio. Overall, Polivy seems positioned to bring new patient benefits, not just in diffuse large B-cell lymphoma, but in other lymphoma settings as well. Next slide. Now let's turn to glofitamab, which is 1 of 2 CD20/CD3 bispecific antibodies that we're developing. And in a moment, I'll remind you of the rationale for developing 2 bispecifics. But first, let's take a look at some illustrative data. So you will recall glofitamab was engineered to have what we call bivalent CD20 binding. So that's 2 binding sites on 1 arm of the antibody against CD20, and then monovalent CD3 binding, so 1 binding site against CD3 on the T cells. And this feature may allow enhanced efficacy against malignum B cells, but in some settings it may also increase the need for hospitalization to manage the cytokine release syndrome, which, of course, is an on-target toxicity of these types of antibodies. But in heavily pretreated non-Hodgkin's lymphoma patients, both aggressive and indolent disease, we are indeed seeing robust and durable response rates with the step-up dosing approach that we'll be using moving forward. And also, as Teresa mentioned, the safety profile with this step-up dosing approach seems manageable. Thus far, most CRS events have been combined -- confined to cycle 1. And so this is certainly encouraging. And therefore, we have a comprehensive development program for glofitamab, starting an aggressive non-Hodgkin lymphoma settings, where cures are possible in principle. And therefore, efficacy receives particular emphasis. And so monotherapy results in the third line or greater diffuse large cell lymphoma setting should readout next year. And we also have a number of combination studies, which include our Phase III trial with GemOx chemotherapy in the second line or greater DLBCL setting. We have combinations with Gazyva or R-CHOP and Pola R-CHP also in diffuse large cell lymphoma. But the big picture here is that we'll always follow the data. And toward that end, we're looking at glofitamab monotherapy in various combinations in indolent lymphomas as well. So we'll see where those results might take us. But certainly, a very promising outlook for glofitamab. Next slide. All right. Now let's look at most mosunetuzumab, which is our other CD20/CD3 bispecific antibody. So in contrast to the 2:1 design of glofitamab, mosunetuzumab has a traditional 1:1 antibody design, which means that each arm of the antibody is monovalent. And this design, as you heard from Teresa, has enabled good efficacy from fixed duration treatment in several settings, with a safety profile that enables outpatient administration. So our mosunetuzumab development plan starts with follicular lymphoma and the monotherapy data in the third line or greater setting there should report out later this year. And here, our combination studies include an ongoing Phase III trial of mosunetuzumab plus lenalidomide in second line or greater follicular lymphoma. We have various combinations in both follicular lymphoma and diffuse large cell lymphoma, including mosunetuzumab and Polivy in the second-line or greater setting. And here, too, we're following the data to determine where the efficacy safety profile of mosunetuzumab versus glofitamab might best serve the needs of patients. And so the big picture here strategically is that this dual strategy with bispecific antibodies could eventually help enable a personalized health care approach in B-cell malignancies as a whole because together, mosunetuzumab and glofitamab may provide optionality across the spectrum of disease severity and the full range of treatment settings in which patients may find themselves. Okay, next slide. So now we'll shift gears to Venclexta. And as I mentioned earlier, this is our BCL-2 inhibitor, which has already brought significant patient benefit in CLL and now a subset of acute myelogenous leukemia. And is now showing efficacy in additional AML context and in myelodysplastic syndrome. So for example, the data here is from a combination of Venclexta and azacitidine in high-risk myelodysplastic syndrome patients. You can see that the overall response rate is 78%. The median overall survival is more than 27 months. Both of these endpoints appear superior to historical treatment outcomes. And based on these data, we initiated the phase trial in frontline myelodysplastic syndrome back in the fourth quarter of last year. And in addition, our ongoing development plan includes the Phase III VIALE-M study in first-line fit acute myelogenous leukemia, the maintenance setting. This will be done together with the azacitidine molecule that was recently developed by BMS. We also have the HOVON induction and consolidation study in newly diagnosed AML patients. And meanwhile, we also have the Phase III CristaLLo study in first-line fit CLL that's ongoing, and it has a minimal residual disease, primary endpoint, and that readout should happen in the 2023 time frame. And finally, we have our Phase III CANOVA trial in multiple myeloma patients -- actually the subset of multiple myeloma patients that have the t(11;14) translocation. That's also ongoing. We expect results next year for that. You may recall that this translocation is associated with an increase in BCL-2 expression relative to another anti-cell death protein MCL1. And the bottom line is that provides a specific biological rationale for the mechanism of action of Venclexta in that setting. Next slide. So speaking of multiple myeloma, which would be a new indication for us, we also have a first-in-class bispecific antibody that we're developing for this indication. And in addition to the CD3 on T cells, cevostamab targets a protein called FcRH5, which is expressed on all myeloma cells. And our dose escalation approach has identified a step-up dosing mechanism, which is producing very good response rates in highly refractory patients and rates of cytokine release syndrome that appear manageable. Most of these have been grade 1 and 2 and we've only seen 1 patient with grade 3 thus far. And so we expect to provide updated results on our Phase I cevostamab program later this year. Next slide. Okay. Now let's take a look at some of our solid tumor programs, starting with ER-positive breast cancer and giredestrant, which is our oral SERD. So we believe that giredestrant has best-in-class potential. And there are several reasons why we believe this. One reason is its potency. You've already heard in vitro studies suggest that giredestrant is 7- to 15-fold more potent than competitor oral SERD molecules. And this was nicely described in a recent editorial as Teresa mentioned. Another reason we believe giredestrant may be best-in-class is its mechanism of action. And so obviously, it's an oral SERD. But in addition to that, our dear colleagues published a paper [indiscernible] in 2019 that showed that when giredestrant binds to the estrogen receptor, it becomes much more difficult for the estrogen receptor to access chromatin. And so that causes profound suppression of ER target genes prior to degradation of the ER protein. And this is important because aberrant transcriptional activation of downstream effector genes, that's the primary mechanism through which the estrogen receptor acts as a breast cancer oncogene. So this is a differentiated mechanism of action with giredestrant. Now a third reason why we think giredestrant could be best-in-class is its safety profile. At the 30-milligram dose, which we're using for all of our registration studies, monotherapy and combination studies, which are shown here. So we've only occasionally seen, for example, any bradycardia, which has been viewed as an on-mechanism effect, it's only been grade 1 bradycardia. It's not been clinically impactful. We've not had discontinuation. We've not had dose interruption. And we've seen no adverse pharmacokinetic effects or drug-drug interaction issues when giredestrant has been combined with other molecules, for example, palbociclib. And that's different than other SERDs in development. For example, the dose of [indiscernible] needs to be cut in half when that SERD is combined with palbo. And by the way, the palbo exposure is also reduced by nearly a quarter in that context. So those are several reasons why we think giredestrant can be best in class. Next slide. But perhaps the most important reason why we think it's best-in-class is the clinical efficacy that we've seen thus far in both the metastatic and early diseases. At ASCO earlier this year, we presented some initial neoadjuvant data shown here on the left, in which giredestrant achieved a 78% mean reduction in tumor Ki67 levels, and 55% of those tumors showed evidence of complete cell cycle arrest after 2 weeks of treatment. So we'll present additional neoadjuvant data for giredestrant at ESMO in just a few days. So be sure to take a look at that. But the totality of the data fueled our decision to initiate an adjuvant study of giredestrant, 30 milligrams a day compared head-to-head to aromatase inhibitors in high-risk early ER-positive breast cancer. And that 30-milligram per day dose is also the dose we're studying in the metastatic setting. We've seen good efficacy across the board in that setting. That includes patients with ESR1 mutations, that includes combination with palbociclib. And as I mentioned, both of those medicines are administered at full doses, no drug-drug interaction issues observed thus far. And so the result of our Phase II study in the second-line and third-line ER-positive breast cancer setting will readout next year. Recruitment for the study is going very well. And we also have a Phase III trial ongoing of giredestrant and palbociclib in first-line metastatic ER-positive breast cancer. Next slide. Let me also just remind you briefly that we have what we believe could become a best-in-class PI3 kinase inhibitor. So GDC-0077 is highly selective for the alpha subunit of PI3 kinase. It can induce degradation of oncogenic PI3 kinase mutant proteins. And it combined with other targeted therapies at full doses. And so an example of that is shown here of the clinical efficacy of our PI3 kinase inhibitor, combined with palbociclib and letrozole in PIK3CA-mutant breast cancer, again full doses. And so we have a Phase III trial of this combination ongoing in frontline metastatic breast cancers that harbor PIK3CA mutations. Next slide. All right. So another important area of emphasis for us involves deploying our pipeline in earlier cancer treatment settings as the science dictates. And Teresa and Bill both mentioned earlier that this is the setting where durable control and cures are most often possible. And also, we believe that we can increasingly leverage emerging liquid biopsy platforms to identify patients who might benefit from earlier treatment intervention. So our adjuvant program spans several solid tumor types. And in 2022, we should have multiple data readouts, which include studies of Tecentriq in neoadjuvant non-small cell lung cancer, adjuvant renal cell carcinoma, adjuvant head and neck squamous cell carcinoma and adjuvant hepatocellular carcinoma, and that's a study of Tecentriq plus Avastin. And in addition, our study of alectinib, our ALK inhibitor should read out in the adjuvant ALK-positive lung cancer setting next year. Next slide. So of course, you've already heard and you already know that one key study from our adjuvant program has read out positive this year. And the relevant data is shown here. In PD-L1 positive patients, we saw a greater than 7-month increase in median disease-free survival with Tecentriq compared to best supportive care. The overall survival data is premature, though perhaps it hence had a favorable trend. We'll have another OS readout next year. So these data have been filed through RTOR and Project Orbis and granted priority review by the FDA. Next slide. So we also continue to explore immuno-oncology solid tumor opportunities beyond Tecentriq. So for example, we're combining tiragolumab, which is our TIGIT inhibitor, together with Tecentriq in several solid tumor settings. Just to remind you, TIGIT is another immune checkpoint receptor with biological similarities to PD-1. And this program was galvanized by the intriguing randomized Phase II data that we saw in our SKYSCRAPER trial in which tiragolumab, together with Tecentriq showed increased response rates and prolonged PFS in PD-L1 high non-small cell lung cancer. And in 2022, we will know a lot more about the potential for tiragolumab. We have 4 trial readouts next year and those include a Phase III trial in PD-L1 high non-small cell lung cancer. That one will replicate the SKYSCRAPER result. We also have readouts in first-line extensive stage small cell lung cancer and in frontline esophageal cancer. And then in addition, we'll see Phase II results in cervical cancer. Next slide. We've also been leveraging our bispecific antibody platform to develop assets capable of inhibiting 2 immune checkpoint receptors simultaneously on T cells. And one of these bispecifics is a dual inhibitor of PD-1 and LAG3 This is interesting because combined PD-1 and LAG3 inhibition was recently validated clinically in melanoma. And that was presented earlier this year. And the second bispecific is a dual inhibitor of PD-1 and TIM-3. And the guiding hypothesis behind both of these bispecifics is that they may help overcome T-cell exhaustion mechanisms and thereby boost the antitumor activity of CD8+ T cells relative to PD-1 inhibition alone. So we're exploring both of these bispecifics across several tumor types, including melanoma, non-small cell lung cancer and esophageal cancer. Next slide. So I'll conclude with a few words on 2 of our benign hematology programs, starting with our gene therapy asset for hemophilia A from Spark. And to remind you, this is an adeno-associated viral vector in which a code optimized factor VIII gene is under the control of a liver-specific promoter. And in our clinical data, in 15 out of 17 participants in the study and at a 2.8-year medium follow-up, we've seen sustained factor VIII expression, durable protein activity. And in terms of clinical endpoint, this corresponds to a 91% reduction in annualized bleeding rate and a 97% reduction in annualized infusion rate compared to the year prior to vector administration. And so the safety profile continues to be acceptable at the dose of studies. But of course, for us, a key goal is to ensure that both the dose and the immunomodulator regimens are optimized. And with that said, we expect to make a decision on the start of our Phase III study of this gene therapy asset later this year. Next slide. Okay. So lastly, we have an anti-C5 antibody called crovalimab. We're studying this in several complement-mediated disorders. And crovalimab was engineered using recycling technology developed at Chugai, which enables prolonged neutralization of the C5 protein when given subcutaneously every 4 weeks. And so Phase III data for crovalimab in Chinese patients with paroxysmal nocturnal hemoglobinuria will readout next year, as I mentioned earlier. But we're also exploring the benefit of crovalimab in other disease settings where the pathophysiology is driven at least in part by aberrant complement activation but these include atypical hemolytic uremic syndrome. There, we're initiating Phase III trials in both adult and pediatric populations. And also sickle cell disease. There's a role for complement there as well and the damage of the sickle cells. We expect to start Phase I and Phase II trials in acute and chronic sickle cell disease later this year. So hopefully, I've convinced you that our oncology pipeline remains one of the most robust, if not the most robust in the industry. And with that, I'll stop, and I'll turn it over to Paulo Fontoura, who's our Global Head of Neuroscience as well as Immunology, Ophthalmology, Infectious Disease, and he'll discuss our Neuroscience portfolio.

Thanks, Levi. And hello, everyone. Very nice to be here again. So I'll take you on a little brief tour of our late-stage pipeline in Neuroscience and Immunology. Talking a little bit about our lead programs in the MS, Alzheimer's, SMA, Duchenne's and Parkinson's and then touching briefly on immunology pipeline as well. Now this is a picture of our current late-stage and early-stage Neuroscience pipeline. And I think one of the important things here is that this is still and remains one of the biggest areas of medical unmet need in the world. And unfortunately, despite that, the number of large pharmaceutical companies developing medicines has reduced in the last decade. But fortunately, Roche has committed to this area for a long time now, and you see the result of that commitment here. This is a very strong pipeline, potentially the best in the industry with several assets, which are either first-in-class or best-in-class targeting major areas of medical need and with 3 launched medicines already, 2 of which last year Enspryng and Evrysdi. And as was mentioned before, if a certain technology or modality makes sense, we will be sure to pursue it. So as you can see here, and you saw a similar slide from Levi in Neuroscience rare diseases we are pursuing several of these technologies and platforms to really target biology at its most fundamental level, because we believe by doing so, we can develop transformative medicines and achieve breakthrough outcomes. Now at the same time, one of the core problems in neuroscience generally has been the ability to measure behavior, to measure endpoints accurately and close to patients. So because of that, for several years now, we've been investing strongly in digital technology and wearables to try and address that. And there's multiple reasons to do this, not only to essentially increase the efficiency of R&D efforts by allowing us to measure behavior more quantitatively and more frequently. Secondly, because actually by measuring these outcomes close to patients, it would demonstrate much better than the value of our medicines and therefore, contribute to create the value proposition for society, which makes sense as well. Now you see here on the right-hand side, just an example of how we're applying these technologies across our whole pipeline in neuroscience. So our ability to measure things like cognition, motor function, speech, et cetera, is being deployed consistently across a range of our trials and programs, and I'll give you a couple of examples of that later on. So starting with Ocrevus and our flagship program in multiple sclerosis. As Teresa mentioned, this remains the only medicine that has been approved for relapsing and progressive forms of MS. It is -- it has the most robust data set ever in demonstrating effects both on measures of relapses as well as progression. And this is really the large remaining medical need for patients with MS is the ability to control long-term visibility and therefore, to control progression. So it's really with that in mind that we've decided to expand the OCREVUS program to study higher doses of OCREVUS. And we currently have 2 Phase III trials, both in relapsing as well as primary progressive patients looking at that and really trying to measure outcomes on the progression of these patients. Now the rationale behind this study is actually a really great example of following the science. Looking at our Phase III data set, which got the approval for OCREVUS, it became clear that at deeper levels of B-cell depletion and higher levels of exposure to OCREVUS, we would see a better benefit for patients in terms of CDP outcomes. And therefore, we're really studying now essentially 1,200 milligrams or 1,800 milligrams for heavier patients to try to maximize the impact on those progression outcomes. And as I said, these 2 Phase III trials have been started in 2020 and are recruiting very fast. Now as a part of that, I also wanted to point out the launch of Floodlight, which is our first SaMD device to measure progression outcomes in MS. Now this is based on this digital platform and wearables on a pretty basic cell phone technology that everybody has that sits close to the patient and measures a number of active and passive monitoring tests. And by doing so, create a different value proposition. On one hand, clearly, better quality data in terms of more quantifiable, more frequent, which is allow us to derive better data insights for patients and hopefully even derive new end points. At the same time, because this is close to patients, it would allow us to measure what's called silent progression, meaning health outcomes that typically are not getting measured because patients see their doctors maybe every 3 months, maybe 6 months. And therefore, this will detect those changes and allow for patients to be better informed in managing their own care. And at the same time, for society, as well, if our objective is to prevent long-term disability, and we know this is the most determinant of health outcomes for these patients. The earlier intervention coupled with early detection of silent progression will allow us to do that much better. So all in all, we're very excited about this and the potential of this has to change how patients are managed. Now switching over to talk about our BTK inhibitor program in MS. This has been a target that has generated a lot of attention in MS. Of course, it builds on the B-cell mode of action story that we pioneered with OCREVUS, but it adds something to it because of the dual mode of action. We know that our BTK fenebrutinib targets both B cells as well as myeloid cells. Now when you talk about BTK drugs, there are several in development right now. So it's really about the differentiation. We do believe this is potentially the best-in-class drug out there. Based on one hand on being the only reversible and highly selective BTK inhibitor. And you see data here on the right-hand side that shows essentially a kind of selectivity assay which demonstrates about 130% plus higher selectivity compared to other BTK inhibitors in MS right now. And at the same time, because it is reversible, it does have a long residence time as bound to BTK, which means that it essentially acts as a irreversible inhibitor from a physiological standpoint, but when you stop dosing, 24 hours later, the enzyme is free again. Now this is a program that we started and has started to recruit this year. We have 2 trials ongoing, both relapsing as well as primary progressive patients. These are really innovative trials. Again, our outcomes here, we're mainly trying to increase effects on progression in these patients. And therefore, we are comparing to standard of care to teriflunomide in relapsing disease or to OCREVUS in primary progressive. And our key endpoint is really what's called a composite confirmed disability progression at 12 weeks. So this is a measure not just of EDSS, but also of upper limb function and walking ability for these patients. Now switching over to Alzheimer's. And this is really an area that has had a lot of interesting news in recent times. I just wanted to remind you all that we are still committed to developing gantenerumab. This is we feel a really best-in-class molecule profile. It's a very selective anti-amyloid IgG1 antibody that targets specifically neurotoxic oligomers and plaques. It has been studied over a decade at very different dose levels, including familial Alzheimer's and it demonstrated very strong biomarker engagement for this molecule, as you can see here in the middle. Essentially at 3 years, over 80% of patients are amyloid negative. At the same time, in familial Alzheimer's we've demonstrated that, but also impact on downstream markers of neurodegeneration, such as Phospho-Tau or CSF neurofilament light chain. Now when we designed a GRADUATE program, which started a few years ago, and it's just wrapping up now and you will have the readout next year, we really wanted to design the best trials to test the hypothesis that targeting amyloid at sufficiently high levels would result in good clinical outcomes. So we made a very -- a few very conscious choices, one of which was to have everyone regardless of ApoE genotype titrated up to the highest dose level, so that's 1,020-milligram subcu every month. So everyone is getting up to that very high level. We've extended the duration of clinical follow-up to 27 months to -- because Alzheimer's is a disease of years and decades, not a disease of month. So we really want to know what the long-term safety and benefits profile is. We have one simple titration regimen that last 9 months, and that's really to be able to bring up safely all the patients up to that maximum dose. And again, we've designed these 2 very large, over 1,000 patient studies that are essentially going to, we think, provide the best data set next year to evaluate the efficacy and the safety profile of an anti-amyloid targeting therapy in Alzheimer's patients. Now we have a couple of follow-on programs I'd like to touch on, one of which is our brain shuttle platform, gantenerumab. This is essentially to get around one of the biggest difficulties in CNS drug development, which is getting enough monoclonal antibodies or biologics generally across the blood-brain barrier and into the target organ. So in this case, we're coupling gantenerumab to our transferring receptor 1 branch shuttle module. And this allows, as you can see here in the middle, in our preclinical animal models, both in mice and in monkeys, to essentially go up 30- to 50-fold higher brain targeting for gantenerumab. Now we already have this molecule in humans, and I'm showing you here on the right-hand side, Phase I data, PK/PD data in healthy volunteers, which shows about an 8-fold higher CSF concentration of the brain shuttle version compared to normal gantenerumab. And we're running right now a Phase Ib trial in about 120 patients with mild to moderate Alzheimer's to look at impact on biomarkers. The other key target in neurodegeneration is clearly TAU. And we have 2 programs in Phase II right now, semorinemab and bepranemab. These target different epitopes in the TAU molecule. And the proposed mode of action here is that you would essentially interrupt cell-to-cell spreading of TAU pathology and therefore, block the second hallmark pathology in Alzheimer's disease. Now in our first trials, and I'm showing you here some of the biomarker data, we did show pharmacodynamic engagement in these patients. We did have 2 trials, one of which called Tauriel readout negatively in prodromal to mild patients. But just recently, we've announced the top line data from our Phase II trial in mild to moderate Alzheimer's patients called Lauriet, in which for the first time for a TAU antibody, we've demonstrated a pretty robust significant impact on cognitive decline in these patients. So we're excited about this data, and it's something that we want to continue to look at. Switching over to Evrysdi. You heard from Teresa, this is the first and only oral splicing modifier that has been developed for essentially the broadest patient population of type 1, 2 and 3 patients. It has really changed the lives of many, many people. It is a very well tolerated, very efficacious drug. And of course, we want to continue to expand on that story. And therefore, we are continuing to study this molecule. And here, I'm showing you data from our RAINBOWFISH, which is our study in presymptomatic babies with SMA. So these are babies with a mutation in either 2 or 3 copies of the SMN2 gene. Now of course, earlier intervention usually leads to better outcomes in these types of diseases. And as you can see here on the right-hand side, we essentially are bringing these babies up to a normal development profile. All of them are sitting up. Some of them are standing up. Their CHOP-INTEND scores that you can see here are essentially maxing out to a normal level. We feel really, really excited about these data. And hopefully, we'll be having conversations with health authorities to be able to expand that label. The other program we're very excited about is our gene therapy partnership with Sarepta for Duchenne's muscular dystrophy. Now of course, Duchenne's is an area of tremendous medical need, and there's really no good disease-modifying therapies out there. So a gene therapy could actually be the first breakthrough for these boys. The microdystrophin gene therapy, we're developing here, has unique properties. The vector is the AAVrh74 vector, which allows for much better tropism for skeletal and cardiac muscles and also avoid preexisting immune responses. The expression of MHCK7 actually has a promoter that maximizes the expression of this microdystrophin gene in cardiac and skeletal muscle cells. And as you can see here, this is data from our 103 study, in which we're using the first time the commercial formulation of this gene therapy in about 20 boys that were dosed with it with 1.33x 10 to the 14 viral genome. So this is our target dose. And you can see here a clear demonstration not just of safety, but also very high levels of expression, both in terms of number of viral genomes per cell, which was almost 4, but also in the numbers of the expression of this protein measured by Western Blot, by immunofluorescence. So essentially a really clear demonstration of the biological efficacy of these gene therapy in restoring dystrophin to these muscle cells. And we're on track to start our first Phase III trials later on this year. Continuing to talk about neurodegeneration. And you've probably seen this data a few times. We just wanted to mention that we continue to develop prasinezumab for Parkinson's disease. This is the most advanced disease-modifying therapy for Parkinson's right now targeting alpha-synuclein, which is one of the key targets in terms of misfolded proteins for Parkinson's. And we had a Phase II trial called PASADENA that you can see the data here for, although it missed the primary endpoint, it was the first time that an effect was seen on the MDS UPDRS part III scale that measures motor symptoms in these patients, both in terms in the change in this continuous variable as well as the change in time to worsening so how fast patients are progressing. Now interestingly, at the same time, we had deployed a digital score based on the Floodlight platform which showed very, very concurrent data. And actually, it was the concurrence of both clinical data and digital data that encourages to continue to study this medicine. And right now, we have a Phase IIb trial called PADOVA that's going to look at slightly more advanced patients which are being treated with symptomatic therapy such as L-DOPA. Now this digital endpoint, just to continue to build on that story of needing to measure more accurately the things that actually matter. As you can imagine for Parkinson's patients that have a variety of motor symptoms on a daily basis, which fluctuate, there are good days, there are bad days, having the ability to measure symptoms really closely and very frequently in these patients could actually be much more sensitive. And that is basically what we are seeing here. As you see on the left-hand side, there's a variety of both daily active tests such as measuring tremor or rigidity, finger tapping, but also passive monitoring of walking, balance, et cetera, which altogether derive a motor score, which has a very high degree of correlation with the MDS UPDRS scores. So I think this is just a great example of how technology can help solve clinical research questions, and we're currently in dialogue constantly with health authorities globally to see if endpoints such as these might actually become acceptable for actually using in clinical trials. Now switching over to Enspryng in myasthenia gravis. As you know, we got approved Enspryng last year for neuromyelitis optica. Myasthenia gravis is another chronic autoimmune rare disease, although there's almost about 1 million patients globally, of which about 80% really don't have stable management of their disease. And then about 10% of them, this can actually result in myasthenic crisis, which can be life-threatening. And right now, standard of care is either treated treatment with steroids or IVIG or then very highly potent immunosuppressive drug would not such a great safety profile. So we actually feel that Enspryng might be a really interesting new mode of action for these patients that would allow essentially safe targeting one of the key pathophysiologies here, but at the same time, with a very convenient subcu dosing every 4 weeks. Now finally, switching over to immunology. I wanted to mention our Gazyva program in lupus nephritis, membranous nephropathy and systemic lupus. This really builds on our long-standing knowledge of B-cell depletion as one of the key mechanisms in autoimmune diseases. This is using Gazyva, which is a glycoengineered anti-CD20 antibody, which much degree -- much higher potency. And we do know that resident B cells in tissues such as the kidney for lupus, is one of the key drivers of the biology for this disease. So we're basing ourselves as well on Phase II data from our NOBILITY trial, which you see here in the middle, which clearly showed compared to a placebo, a much higher degree of complete renal response at week 76, about a 50% response in people who are sustainably depleted B cells. And therefore, this could really be a breakthrough for these patients. And we're continuing to study this molecule now in Phase III trials in lupus nephritis, which we started just about now, but also in membranous nephropathy and in systemic lupus. The final molecule I wanted to mention is our program in fibrotic diseases we've been traction to in IPF. Now this is still very much a major medical need. Most patients with IPF die within 5 years. Standard of care hasn't really changed much. There are 2 drugs, which are approved but don't really change the progression significantly. And we're excited about this molecule because on one hand, the mode of action is that it changes fibrosis differentiation from pro-fibrotic, pro-inflammatory to a proresoltive state. And at the same time, we have Phase II results that show that on top of standard of care in these patients, for the first time, we've demonstrated an additional benefit, so a substantial reduction in the progression of loss in FVC at 6 months. But at the same time, sustained endurance in terms of the 6-minute walking test. So this could really be the first breakthrough for these patients. And therefore, we're very, very excited to have just started our first trial in IPF this year. So with that, I'll thank you and turn it over to Nilesh.

Thank you, Carlo. My name is Nilesh Meta, and it's my privilege to share an overview of the Ophtha portfolio at Roche. Over the next few slides, I'll cover the landscape and unmet need in retina. I'll overview data from our late-stage assets faricimab and port delivery system, and then touch on the Ophtha pipeline and our personalized health care approaches. So as you can see here, in spite of the tremendous impact of IVT anti-VEGF therapies, there's still a high unmet need in retina. In clinical trials where patients are dosed frequently, approximately 3 to 4 out of 10 patients don't achieve 20/40 vision. And that's usually required for driving. But even more so in the real world, patients are not able to adhere to their treatment intervals either because health care systems cannot provide sufficient capacity for injections or patients can't manage the frequent visits, and as a result, lose vision. Doctors will tell us that approximately 50% of patients can eventually be extended to Q12-week dosing. But with both faricimab and port delivery system, we believe we can provide physicians with truly innovative options to address this unmet need for durability and thereby improve real-world outcomes. The key message here is that durability is directly linked to vision outcomes and the benefits that gives to society. So firstly, let's touch on faricimab. It's been filed in the U.S. and in EU, and we've got approval for this expected in 2022. Faricimab is the first bispecific antibody developed specifically for ocular administration. It simultaneously binds to 2 distinct pathways well-known in retinal pathology, so the well-known anti-VEGF part, but also now to angiopoietin-2, which has been implicated in vessel destabilization. As Teresa mentioned, it's the first new MOA in retina for over 15 years and really distinctly different to monotherapy anti-VEGFs. It's engineered with a Modified Fc portion that allows for longer action in the eye and but faster clearance systemically. So quite unusually, faricimab was studied in both AMD and DME simultaneously. It's the first time the company has conducted these studies in parallel. And both indications have been filed with key regulators around the world. Further, we have Phase III studies ongoing in RVO, which is a third major retinal indication. Let's touch a little bit on the data here. The positive data in DME and neovascular AMD has been presented at conferences through the year. On the left, you'll note the consistent data across all 4 Phase III studies in both indications, showing approximately 75% to 80% of patients being able to be extended to Q12 and about half to Q16. This is the first time this level of durability has been demonstrated in a Phase III setting. All of these arms in the faricimab trials provided BCVA gains similar to faricimab dosed every 8 weeks. In DME, the anatomic outcomes favored faricimab. So this has been both absence of DME and absence of intraretinal fluid. The safety profile was in line with current anti-VEGF IVTs, with small numbers of adverse events and no cases of vasculitis reported. And in our opinion, this sets faricimab to be a potential future new standard of care in the IVT setting. Furthermore, we will be delivering on long-term data so RHONE-X and AVONELLE-X, which are ongoing and will provide 4 years of longitudinal data for faricimab. So shifting to another innovative program, the port delivery system with ranibizumab. So this has been filed in the U.S. and in the EU in quarter 2 this year, and FDA approval is expected in October of this year as recently announced. Just as a reminder, the port delivery system is a permanent intraocular implant about the size of a grain of rice, and almost all patients were able to be dosed every 6 months. The implant procedure is an approximately 30-minute outpatient procedure and the refills are akin to an IVT injection performed in an office setting. I'd like to point out here the Pagoda study, which is also now fully enrolled. It's about 540 patients. And the speed of the study enrollment really suggests both the importance and the interest of retinal specialists in the PDS. I should note that there is robust IP protection around the PDS device and the formulation that ranibizumab that goes into it. Touching on the data. In the Phase III Archway study, PDS patients dosed every 6 months showed equivalent visual acuity gains to monthly anti-VEGFs. And nothing really has come even close to this level of duration with almost all patients going every 6 months with no drop in BCVA. 98% of patients did not require a supplemental treatment. And further, in a survey conducted as part of the trial, 93% of patients reported a preference for PDS over IVTs. The PDS implant and refills were well tolerated with a safety -- with a favorable profile. So as these programs are in regulatory review, we're planning for global launches. I think Teresa mentioned that this is the first time we're going to have a global footprint in Ophthalmology. And for the launches, we're really focused on ensuring consistency of the implant and safety for patients. To support these objectives, we'll be supporting eye surgeons to practice both the implant and the refill proceed using virtual reality-based systems. And in addition, we'll have field-based surgical device sleeves in place to help surgeons in the OR. Hiring and training are ongoing. We know how important payer discussions, access and reimbursement are for this community, and these are progressing as we speak. So further to the late-stage assets that I've just shared some data on faricimab and port delivery system, here you can see the robust pipeline that we have in Ophthalmology. There are sort of 8 new molecular entities that are in either Phase 1 or 2. And this really forms one of the leading pipelines in Ophthalmology. I think as the science on different pathways involved in pathologic eye conditions, and the associated imaging and clinical biomarkers continue to evolve, it makes for a very exciting time to be in Ophthalmology research. One of these platforms that I really want to highlight is DutaFabs, I think Bill mentioned it earlier, which really next generation bispecifics. They can be concentrated significantly and are being developed to be compatible with port delivery system. The next one to progress into clinic is anti-VEGF, Ang-2 bispecific but other pathways are also being assessed. Finally, beyond the therapeutics, we believe there are other technologies. I think Paulo has alluded to them in NS as well, but that can support patients and clinics to have a better experience of managing their chronic eye conditions, and therefore, hopefully, better personalized care for patients and their families. One to call out here would be myVisionTrack, which is an app-based vision test, which can provide patients and their physicians confidence to monitor vision between extended visits to the office. The teams in ophthalmology continue to work on imaging modalities and have a number of real-world data collaborations ongoing, which you can see on the slide here. With that, it's my pleasure to hand over to Barry Clinch.

Thanks, Nilesh. Next slide, please. So I'm grateful to have the opportunity this afternoon to speak to you a little bit about our Infectious Disease franchise. I'm going to talk in 3 sections. Basically, first of all, a little bit of time spent on our HBV franchise, and I'll talk about the different molecules we have there. The bulk of the time I'll spend on our SARS-CoV-2 franchise and looking molecule by molecule at the 3 key components of that. And finally, to finish up with little bit of an outlook of how we're thinking about where the pandemic and this virus are going to go over the next months and perhaps even years. So we can move to the next slide. So in our hepatitis B portfolio, here, I think it's a good illustration of some of the points that Bill had made right at the out session and others of my colleagues have to. We're interested in a cure treatments so the goal here is a functional cure for these patients and a wide modality base as well. We're looking at both siRNA approaches, locked nucleic acid, small molecules. So we're taking a technology -- technological approaches that make sense to us given what we know about the disease biology of chronic HBV infection. And we believe that we will need different approaches to deal with this virus. Part of it on the left-hand side of this slide is directly attacking the virus itself. That's going to be important. But on the other side, we also believe stimulating the immune system to gain control of the virus ultimately is going to be important as well. So we have MOAs to target both sides of this. And we believe combinations will be necessary. Our thesis is that it will be something that targets the virus plus something that modulates the immune system, either the innate or the adaptive that's going to deliver the outcomes that we want for patients in terms of a functional cure. If we can go to the next slide, a key part of this is going to be rapidly evaluating these different combinations. As you can imagine, with complex biology and a number of opportunities, there are plenty of permutations and combinations that can be tested. And so we've got this now running this Phase II platform study, which will basically allow us to seamlessly take in and out different combinations from our pipeline to study them. And given the ambition that we have for functional cure and the fact that we know what we need to measure, so we're looking for S antigen loss because we know this is a key driver of the disease stage of chronicity for these patients. This allows us to actually have relatively small and efficient studies that will provide very robust data outputs that will allow us to select between these different combinations as well as that, it will also allow us, as you can see on the left-hand side, the opportunity to take interim looks at different points. We're going in with certain assumptions about how long we need to treat and what effect size we'll see, but we can actually monitor this relatively real time and make adjustments as we're going along with these interim analysis. And that may allow us to find surrogate markers of control earlier on or adjust, in fact, the duration of treatment that's ultimately going to be required to deliver a functional cure for these patients. On the right-hand side there, you can see some of the combinations that we have running at the moment. But this is ongoing at the moment, and we expect to see data readouts in the coming months from this that will start to allow us to home in on that ultimate combination. So we can jump to the next slide and switching gears to SARS-CoV-2 and our response to the pandemic. So before going into the molecules in particular. I just want to take a step back and have a look at the totality of what Roche has done in the pandemic space over the last 12 months or so. And this comes from both the diagnostic side and the pharma side. And on the diagnostics side, our colleagues were very rapidly able to bring forward not only typical hospital analyzer-based gold standard PCR testing, but also near patient antigen testing as well as antibody testing to verify whether patients had already being exposed to the virus. So we are testing from all the necessary angles to help with both the diagnosis and the epidemiological management and characterization of this virus from a very early stage. You can see much of this happened early in last year. So it was extraordinarily rapid development. On the treatment side, we've looked across -- and I'll go into these in more detail, we've had 3 key approaches through Actemra in immunomodulation, Ronapreve as a direct antiviral but a large molecule and AT-527, again, direct antiviral but a small molecule. And various readouts have been going on throughout this year, but I'll talk about them in a little more detail as we go through each molecule. So we go to the next slide. Touching on our strategy, perhaps before I get into the molecules themselves, we were obviously faced with the same task that everybody was last year of how do we address the pandemic. And one of the things we wanted to do was have an impact and play to our strengths. And the pyramid or triangle on the left-hand side of this was our way of looking at the landscape and thinking about where could we have the biggest impact and where would best play to our strength. And while I think we all realized that vaccines when it comes to infectious disease are the best, prevention is always better than cure. For us, getting into the treatment space was definitely going to play to our strengths and allow us to have the biggest impact on the pandemic. And we sort of stacked these up, starting at the bottom with an oral antiviral because of the breadth of impact that could have, the number of patients that would have the opportunity to treat, followed by neutralizing antibodies. And for those patients in hospital, immunomodulation was clearly important. And so we have AT-527. We have Ronapreve and we have Actemra. And although I won't go into it in detail, you can see on the right-hand side, the degree of activity and a degree of success we've had in bringing these molecules to patients. We can jump to the next slide. Looking now molecule by molecule. So this is looking at Actemra specifically to begin with as an IL-6 inhibitor. It's already been widely approved in a number of indications before now. And it was clear from relatively early on in the pandemic that IL-6 seems to play an important role as a driver of inflammation, especially in the critically ill patients. And so it looked fairly obvious that Actemra could be useful, but it needed data to back it up. Roche has conducted a number of studies. You can see on the right-hand side, EMPACTA, COVACTA, REMDACTA, but also some of the big platform studies also saw the merit of Actemra, brought that into their platforms, namely RECOVERY and REMAP-CAP. And importantly, the final analysis of all of this data at this day clearly shows us that there is a place in the hospitalized patients for treating COVID-19 with Actemra. Both the RECOVERY study very clearly showed it with a 14% improvement even on the back of steroids in the risk of death. And then more recently, with a large meta-analysis of all of the Actemra data including more than 8,000 patients, underscoring that particular benefit and leading to WHO guidelines for the use of Actemra. So I think a great success story for this molecule. Moving to the next slide. Switching directions on now going from immunomodulation to a direct-acting antiviral. This is the collaboration with Regeneron and their cocktail of 2 distinct neutralizing antibodies targeting the spike protein on the surface but in nonoverlapping, noncompetitive spaces. And we thought this was from the outset, particularly important with our experience with respiratory viruses and knowing their ability to evade treatments. Having dual antibodies targeting different sites, we believe it's going to be very important for the durability of a treatment like this. And indeed, that has been shown to be the case with even up until today with variants of concern and variants of interest, all well covered by Ronapreve. And here again, multiple studies are being run together with Regeneron through different patient populations going from outpatients who are at risk through prevention of the post-exposure setting and to hospitalized patients and as well, again with the RECOVERY platform, all giving positive readouts, meaning that we've got a very impressive data set for Ronapreve ranging from impact on the risk of hospitalization or death into impacts on the likelihood of having symptomatic illness, and the same with the RECOVERY study showing a 20% reduction in the risk of death for patients who don't -- for seronegative patients who don't amount to their own immune response. So this is a very robust set of results, and we're working very hard to roll this out and make it available to patients around the world. So if we can go to the next slide. And finally, within the COVID space, we have AT-527. This one is certainly the earliest in its progression compared to the previous 2 I've just spoken about. This is a small molecule now, direct acting antiviral that targets the polymerase right at the heart of the machinery of this virus and its ability to make still more viruses. This is a very nice molecule that has the ability to target 2 parts of the polymerase, not only the catalytic site where the new nucleic acid has been synthesized, but another distance site from that, which is also important for the viruses ability to replicate. So we're targeting 2 different parts of the polymerase with this molecule. And although it's early stages yet -- although it's only early stages, we have some very encouraging preliminary data on the viral load impact from the interim results from Atea's own Phase II study in a very mild to moderate hospitalized patient population. It's early days. These are encouraging results. We've got our own studies in a Phase II, which is intended to confirm the dose. And in parallel, we've also started a Phase III trial, which is enrolling currently and where we expect results later this year, but so far, so good. And finally, we'll turn my attention to what might happen next. This is a very difficult challenge. I think all of you will appreciate predicting the future is probably never a very good idea, and I don't think COVID is any exception to that. We've tried to think about this nonetheless. And broadly speaking, we sort of seen this in 3 possible categories. You could have a very optimistic outlook on the left-hand side and saying, no -- hang on a minute. We've got great vaccines. We've got great treatments. This virus is on the way out. In fact, maybe it can be eradicated completely. And this seems very unlikely. This seems unlikely when we thought of this some time ago, but it seems even less likely now with the emergence of the Delta variant. On the other hand, you could be very, very pessimistic indeed and think this virus is just going to outfox us at every step. And no matter what we've done and any success we have, it is only temporary. But I would say also a low likelihood that this is true. The more likely scenario sits somewhere in the middle, where we think that everything that we're doing will be durable, and it will contain and it will manage the virus, but eradication will not be an option. It's going to be something we're going to have to learn to live with. It would be expected to become endemic. It would be expected to become seasonal and join the panoply of winter viruses like RSV, like influenza. And with that sense, that while it will become much, much easier to manage over time, it will still need management. And with that in mind, we're going to continue to develop diagnostics. We're going to continue to develop treatments to manage this important pathogen. At that point, I will finish. Thank you.

Yes. Thanks a lot to all the presenters. And we are now moving into the Q&A. We have 1, 1.5 hours reserved for us. So we are really very much in time. As promised very proud of our presenters. There is really no slippage. And maybe, Henrik, you could kindly share with us what should be done, what can be done in order to address the questions. And maybe you also can give us a hint on how this survey works in the end, what the people have to do and what they also have to do in case they don't want to address these questions.

Yes. Thank you, Karl. So we are coming now to the Q&A session. [Operator Instructions] And with that, I will hand over again to Karl.

Yes. Thanks a lot. I have to admit I couldn't hear all what you said, but I hope that it was transmitted via the phone and the people on the line could actually listen and hear what you said. Maybe just first for this reason, take a question from the chat. It's from Sarita. She was asking how important the head-to-head for oral SERD versus fulvestrant in the H1 '22 so beginning of next year to provide incredibility to the SERD class as an earlier line of treatment. So how important is the trial? And her follow-up question was, would a differentiated safety profile without superiority efficacy be able to support filing, maybe Levi, that's one for you.

Sure. Well, maybe I'll start, and I'll let Charlie add. So A big picture here. So obviously, hopefully, you all appreciated our enthusiasm for our SERD. In general when you think about the study that we're running head-to-head against fulvestrant in later line -- in later line ER-positive breast cancer. One thing that we know as a field is that the later you go in treatment of ER-positive breast cancer, the lower the dependence on the estrogen receptor in those tumors. So when we think about -- we often talk about patient heterogeneity as a buzzword. And ER-positive breast cancer, that's what we're referring to when we think about ER-directed therapy. The later-line metastatic patients are less dependent. There are still some dependents but they're less dependent. So what that means is that we think from the efficacy and the pharmacokinetic data that we've seen for giredestrant, it has a chance to perform very well head-to-head against fulvestrant. And so we'll be looking forward to seeing what those results look like. But it also means that it's difficult to extrapolate what that will mean for earlier lines of treatment, where the dependence on the estrogen receptor is much more profound. So we're looking forward to seeing the results in later line, but I think we have to be cautious about over-interpreting those results in later lines to what they might mean earlier lines, which is a very different biological question, but I'll let Charlie comment to maybe on the other element of the question.

No, I think we've already addressed it. And as you point out, I think the preclinical data and the emerging clinical really do suggest, given its mechanism of action, that giredestrant is -- appears superior the available drugs that inhibit the estrogen receptor pathway. And I think the readout in the coming year in the second, third line setting against fulvestrant will be important. And we're also really looking forward to the data we see at ESMO next week, but I think really speak to the clinical [indiscernible] of the unique mechanism of action that we're seeing with this drug.

Yes. Thanks a lot. So I would say let's go into the question via the phone Vimal from Sanford. I'll open your line now, please.

Great. So maybe just starting with POLARIX. Maybe you've had a little bit more time to digest the data so I mean I guess, where is your conviction on this becoming the new backbone therapy standard of care? And if it is very high, why is this not a global $5 billion opportunity. You mentioned $2 billion at the 2Q call. But if it becomes a standard of care, shouldn't be much larger than that? And then just tied to that, could you just provide us an update on filing? Is it fair to assume auto will be used? And then my second question is maybe just on Paulo on gante. I mean, I appreciate this is a discussion maybe for the future, but I wanted to hear your thoughts on duration of usage? So given plaque reduction continues over multiple years, I'm just curious how Roche -- what Roche believe will be the duration of use should we see a positive cognitive outcome? Could these drugs actually be used for 3, 4 years or longer? And just how much of a role will price play in that duration, i.e., do you think a significantly lower price point versus the current pricing benchmark will drive longer duration?

Okay. I can certainly start, Karl, with the first part of the question on POLARIX and then obviously turn to Teresa and Bill on the commercial opportunity. I think as been said, the results of POLARIX are really important in that it's been 2 decades and more than 11 randomized trials since R-CHOP set the standard of care in diffuse large cell lymphoma, many attempts to unseat R-CHOP with all its benefits until now. And POLARIX shows what we think is clinically meaningful, statistically significant and an important end point of progression-free survival. And we believe that the data which will be presented at an upcoming meeting do support this, representing a new standard of care. So we're excited in terms of the filing. We're anticipating global filing. We really can't speak to the details of that now while it's under development. But we're looking forward to making this important breakthrough available to patients globally. But let me turn to Teresa and Bill on the commercial opportunities.

Yes. Thanks, Charlie. I mean I think we talk about this as a multibillion-dollar opportunity. So I think we're still in the process of really understanding exactly how big this is going to be and who is going to use it and where. I mean I share Charlie's conviction that I think this is practice-changing data. First time in 20 years that we've seen in advance over R-CHOP, which is a highly entrenched regimen. But I think this is going to be exciting data for the community when they see it.

Yes. So Karl, should I take the other one?

Yes, please. Yes.

Yes. No, I think it's a really important question, right? And I go back to the comment I made, which is Alzheimer's is the disease of years and decades and not a disease of months. And therefore, I think long-term therapy is going to be required, right? Now how long can you sustainably have therapy with a subcu formulation. Well, I think convenience actually matters a lot for these patients, so the ability of not having to go to the hospital for complicated IV infusion every month or not having to do a lot of monitoring or monitoring based on genetic blood testing. I think all those things count. The long-term safety and efficacy profile counts a lot. And it's again, why I mentioned that we designed our trials to go over the typical 18 months, more towards the 27 months, and of course, to generate long-term data. I think those data are going to be really, really important to continue to demonstrate the benefit. The long-term safety of these medicines is super important. Therefore, I think for as long as patients are getting benefit and the safety profile is acceptable, and the drug is convenient. Then, of course, other factors play in pricing you mentioned, access, all of that. There's a lot of health care readiness work that needs to get done because these are going to be hundreds of thousands, if not millions of patients. And I'm not minimizing the challenge that, that might pose. But generally, from a medical, biological standpoint, I think long-term therapy is going to be the way to go. And not just in patients with Alzheimer's but also patients who are at risk of Alzheimer's, presymptomatic patients. And we already have numerous trials, ATHENA community out there addressing this population as well. We already know that if you just look at the pathology -- the pathology starts decades before clinical symptoms start. So in the presymptomatic stage, prevention of Alzheimer's is probably going to be a pretty big thing for these types of therapies.

Sachin, you want to be next one, Merrill Lynch.

A few, please. Firstly on faricimab and the commercial opportunity. When discussing the launch, Teresa, you mentioned that physicians like to switch. So wondering if you had some more metrics around that as we think about the first 12 months of launch, i.e., what percentage of this market do you view as poorly controlled or very frequent injectors? How would you define that, that is low-hanging fruit for that first 12-month switch? Second questions are on TIGIT, if I may. And so thank you for outlining the data next year. I just wanted to check on SKYSCRAPER-06, which is your long allcomers, chemo combo for the broader PD-L1 strategy. There have been some debate of some interim data in '22. Is that still on the cards for next year? And then a bigger picture question. You remain, I think, the only player in Phase III outside non-small cell. And that clearly gives you a time advantage. But I wonder if you could speak to the level of confidence in your broader solid tumor strategy given the lack of Phase I/II data into those pivotal reads. And then maybe just one clarification for Teresa on POLARIX when answering the prior question, you talked about how broad this may become. Can you just provide some more color there, i.e., how high risk was the POLARIX population? And is extrapolating that more broader the debate as you frame the commercial opportunity?

Yes. Thanks a lot. Maybe, Nilesh, you take the first one on the ophthalmology.

Sure. Thanks, Sachin, for the question. So for faricimab, we believe really that the profile of the product that we're seeing given that it inhibits both pathologies of disease, providing some vascular stability through the Ang2 and then getting 50% of patients up to Q16 is pretty much a first-line setting. So we think that de novo patients will also benefit from the product. From the most recent data we have in terms of where patients are, the most physicians will tell you about 1/3 of their patients can't really go beyond either Q8 or Q12 and about half of their patients get to Q12 over time. So I think there's a significant opportunity there for faricimab to be used.

Yes. Thank you. Maybe, Charlie you or Levi on the trial design, POLARIX addressable patient population. And maybe you can also kind of take the TIGIT one.

Sure, absolutely. So I think the first question was do we anticipate a readout from SKYSCRAPER-06, which is our first-line trial in patients with non-squamous cell -- non-small cell lung cancer. And the answer is that Phase II readout is anticipated for next year. And we believe it will inform a larger Phase III study in that effort. With regard to your existing bets of trials, as you heard, we have several readouts next year. We think actually the results of CITYSCAPE, the biology of what we know about TIGIT in these diseases, including small cell carcinoma, where the poliovirus receptor is present, we think the data do support these calculated bets. And we're looking forward to all these studies reading out next year. Levi, I don't know if you wanted to add anything.

No. I think, Charlie, you've covered it with the aggregate data to the disease biology data, so that's the TIGIT ligand in non-small cell lung cancer -- the general importance of immunotherapy in small cell lung cancer, that is. And then in general, the aggregate preclinical and early clinical data that we've seen, we think, support the best that we're making.

And if you could you also kind of comment on the POLARIX situation, patient population. Do you -- I mean we have always said that we do expect actually a broad label because the trial setup was also in the broad patient population. I guess, Sachin's question was along these lines. So what do you expect in terms of, let's say -- yes.

Yes. So we anticipate -- we're anticipating a broad label for POLARIX. I mean that's certainly been the precedent that was really set by R-CHOP. So given the clinical benefit of progression-free survival in this population in the first line and potentially current setting,across patients, we're anticipating a broad label in the setting of first-line therapy for large cell lymphoma.

Okay. Thanks a lot. Sachin, I hope this address all your questions. Next one would be Jo Walton from Credit Suisse.

I hope you can hear me. I got 3 questions, please. Firstly, you are investing a huge amount to R&D. You've increased your R&D spend. I wonder if you can tell us how you're thinking about that return on investment. Do you think that things have changed such that you need to invest more to get the same return? Or are your investments in technology and your confidence and satisfaction, we should be able to see at least the same return on the next $10 billion of investment that we see on the first $10 billion? My second question goes back like to the beginning, some discussion about costs and how productive in terms of increasing extra sales, but we're also increasing headcount as one [indiscernible] Could you give us a sense of how far down that journey you are? Do you think that you've got 2/3 of the way down there? Whether you're looking in the next 5 years, you could be making the same level of change. And my final question is just an update, please, on biosimilar erosion. So just if we could have [indiscernible] what your expectations are. You're [indiscernible] very confident. Give us a number which would be amazing to that matching, but just your view of biosolution over the next couple of years.

Thank you. Bill?

Sure. Thanks for the questions, Jo. In terms of the R&D productivity, I think it's probably the most important strategic topic for the whole industry. And I would just say we are not increasing our investment because we think productivity is going down. We actually think -- first off, there's definitely a battle to secure high R&D productivity based on disease unmet need, disease biology, smart clinical trials and things. But we actually hope that we can increase R&D productivity based on the things that I mentioned, both technology opportunities, but also opportunities to be more effective with people's time and people's work. And we have a number of opportunities -- or a number of activities that are going on with that right now. I'll -- maybe I'll see if Levi wants to comment more on that. But let me just finish, in terms -- you asked on the commercial side how far are we down the path in this efficiency journey. I would say we are -- it's hard to say exactly, but maybe halfway. The good news is that we probably -- we've probably done most of the reductions that we need to do. And now we've got a lot of new products coming. And the opportunity we have looking forward is launching new products without necessarily adding lots of people, again, just taking advantage of the great talent we have and the ability of people to do more than what we've asked them to do in the past. Now again, not more hours of work, not more volume, but just handling more complex tasks rather than dividing those jobs up into little pieces and creating new positions that sort of make more complexity, but really benefiting from the full capacity of our people's talents and abilities. In terms of biosimilar erosion, I think we've guided since the beginning of the year, we thought biosimilar erosion this year would be about CHF 4.6 million. And I think we're still thinking that's about that. So...

Yes. Bill, maybe I have another one here, which goes into the same direction, John Murphy, Bloomberg. He was asking if you have any sense on the percentage of Roche's current clinical spend in oncology versus non-oncology. Do you expect now to spend more relative to oncology or to spend relatively less in oncology? Can you give us any kind of an idea here?

Yes. And I'm going to hand off to Levi, if you have additional comments, Levi, on the productivity question. I -- it's hard to say. I mean we follow the science. That's one of our key principles as a company. And if we have more opportunities in oncology than outside of oncology, we'll increase investment. If we have less, we'll decrease. I would say right now that, that ratio is relatively constant. It's not changing a lot because we have a lot of opportunities in oncology, as you've heard about today, and we have a lot of opportunities outside of oncology. So I think maybe one way to think about it -- because sometimes people ask, well, are we really committed to some of these other areas. And I would say, our investment level, while we are the leader in investment in oncology in the world, we're also the leader in investment in a number of other therapy areas. And I think we aim to continue with our plans to bring many more breakthrough medicines in many diseases, including oncology. Levi, other things you want to say?

Yes, absolutely. So maybe I'll just add that Bill in his presentation talked about at a macro level the shifts that allow the increased R&D investments and what that's meant from a workforce standpoint. Within R&D, we similarly see opportunities for shifts to make sure that the work that we're doing and sort of the infrastructure and workforce infrastructure costs can be optimized so that the investment is maximally going to the new clinical trial, new molecule opportunities. And that involves changing ways that we work. Bill mentioned how we leverage technology and platforms and scale. And also the kinds of work, the high -- what's higher versus lower priority work and making sure that we're removing lower priority work. And we've already begun to take steps in that direction. We think there's more that we can do. So the increase in R&D productivity is not just a macro shifting from other parts of the company into R&D. It's also within R&D, optimizing our work and how we deploy our resources.

Yes. Thanks a lot. Next question would be from Tim Anderson.

Okay. Can you hear me okay?

Yes, we can hear you, yes. Thank you.

Okay. Sorry, can you hear me now?

It's perfectly fine. The line is good. We can hear you clearly.

Okay. Great. I have a question on gantenerumab. I'm worried that the risk of failure with the GRADUATE studies might be higher than with monoclonals that rely on IV dosing. It seems like with subcu dosing, you may not get the same pharmacokinetic peaks that you would get with IV, so you get less drug across the blood-brain barrier and a slower rate of plaque reduction. And it seems like we've already seen that slower rate play out [indiscernible] from Marguerite and Scarlet RoAD. So you eventually get deep plaque reduction, but it takes longer to get there. I'm wondering if these trials might time out too early where you don't see a cognitive benefit because you didn't reach deep plaque reduction quick enough. So that's one question. And then on your trial program, mixed results. One trial was a total awash. The other trial showed one endpoint that was positive. But when you describe that program overall, you still seem reasonably excited. So can we confirm that you think that the recent positive data point is strongly supportive of efficacy here? And where do you go with this program? Certainly it's not a registrational data set.

All right. Thank you. Two very, very complex questions. I'll start with the first one on gantenerumab. So there's actually no evidence that you get more drug on target based on IV versus subcu. And even if you do get the peaks -- and during the peaks, you might get more drugs through, what actually drives target engagement in the brain is how the antibody binds to plaque. So it's buildup. It's not -- so what I mean is that the pharmacodynamic effect is not dependent on the pharmacokinetic profile directly. It's depending on the accumulation of drug at target. So we actually have a lot of data on that. We did have the IV form of gantenerumab. We did all the modeling around that. And therefore, I actually am not convinced that that's true. Secondly, when you mentioned about the rate of amyloid removal on our open-label trials, there's 2 points I'd like to make there. The first of which is that, that data that I showed you today and the data from our open-label trials are actually from patients that originally for the first 2 years,were on very low doses. And they were only moved up to higher doses after the futility of Scarlet RoAD. So we're actually not seeing what the real -- let's say, the real dynamics of amyloid removal might be on our GRADUATE program. And again, because this is a very rich data set and we're able to really model these data, and I think we've presented that, what we're predicting is actually roughly about a 25% to 30% reduction in amyloid after 2 years, which is very comparable to, say, other drugs out there. The second point I'd like to make on that, there's actually no -- and I think I mentioned it, there's no clinical correlation between speed of amyloid reduction on PET and clinical outcomes, right? And that's an important point because what we're actually measuring, we're measuring essentially plaque. We're not measuring what happens to oligomers, we're not measuring what happens to the neurons around them. This is basically a good surrogate for pharmacokinetic effects. It's a good surrogate endpoint, at least the FDA seems to think so. But it's not a direct predictor of clinical outcomes. So based on that, I mean I feel very, very confident that, one, we have the right dose. We have the right duration of dosing that the subcu on the contrary will prove to be an advantage because of convenience, because of access and also because we deliberately wanted, as I mentioned, to take time to titrate every one up to the highest dose. We actually design trials that are going to last 1.5 years -- sorry, 2.5 years. So much longer than is typically done. So I guess the bottom line is that I actually am not concerned that the trials are going to read out too early. I think this is still the best design trials out there. Moving over to tau. You're right, there are mixed results, both with our antibodies as well as with antibodies from other companies out there. And in retrospect, I don't think that's remarkable given that this is very new biology. We're trying to tackle what is clearly one of the key pathologies in neurodegeneration. And I think that makes a lot of sense. And whether this is the right modality or the best modality, whether an antisense approach or a gene therapy approach might be better, we don't know yet. What I think is exciting is that for the first time,in any trials in Alzheimer's, we were able to show a real benefit, about a 40% reduction in cognitive decline in mild to moderate patients. This is something that is actually quite exciting because, again, it's the first demonstration of that potential. Does it mean that we're going to rush to Phase III and this is the conclusive proof? No. It's still early days. We're still looking at those data. We're, I guess, excited enough that we still feel that might be advantages in pursuing this type of molecule, but there's a lot of road ahead for us, for sure.

Thank you. Thank you, Tim, for the questions. Simon Baker, Redburn would be the next one.

Just going back firstly to the question of R&D productivity and spend. I noticed that looking at the paper describing the story of the discovery of giredestrant, comparing that with fenebrutinib. On the giredestrant paper, about 1/3 of the orders were from Charles River; whereas fenebrutinib, which was only published 3 years earlier were 100% Genentech employees. So I was wondering if that marks a shift in your approach and utilization of outsourcing and whether that was driven more by capacity or cost. And then a couple of questions I suspect for Teresa. You gave us the chart of oncology visits in the U.S., which showed back to almost normal level. That data was for June. I just wondered if you had any newer data there. And also if there were any meaningful differences by tumor type because I think, previously, we've seen lung and breast have been the most affected. And then finally, on Phesgo, you mentioned the high conversion rate in the U.K. I assume that's driven by price -- everything is driven by price in the U.K. But I just wondered if there are any other factors and what that means for the potential of Phesgo in a world of biosimilar setting.

Yes. Then also for the R&D productivity, Levi, over to you.

Sure. Yes. So I -- so under the specific example that you gave of the giredestrant paper and also versus fenebrutinib. I would say that I wouldn't read too much into that. Those are different molecules and different points in time. And so the nature of the work was different. I wouldn't draw too much in a way of conclusion there. As a general picture, though, we are always looking at where is it -- where does it make the most sense to get the work done that we need to get done and where does that mean investing internally in our workforce and the capabilities that we need versus where does it make more sense to leverage external capabilities. In a rapidly shifting R&D landscape as we're in, we always have to ask these questions. So we're always looking at this, but I wouldn't read too much into the specific examples that you get.

Great. Thanks, Levi. And in terms of the other 2 questions. So related to oncology visits, you're right, in certain parts of the U.S., we are seeing a -- that are harder hit by the Delta variant, we are seeing visits slow down. For the most part, though, we're not seeing too many differences by tumor type. There had been that consistent sort of slower comeback in hematology. Interestingly, these results are pretty consistent with what we're seeing in Europe as well. So mostly back to normal, a little bit of a softness in hematology. And where the Delta variant is flaring, unsurprisingly, you're seeing a little bit of challenge in patients going in for their care. In terms of the U.K., I mean, so -- it's not just about price. I think part of this idea that Bill has been talking about, about cost to society is also what it takes to administer drugs. And part of what made -- has made the Phesgo launch in the U.K. so successful is that it is able to actually decrease the amount of resources the hospital staff need to put towards infusions. And so that's a different kind of savings and a different kind of efficiency in the system that is going to be more attractive to some systems than others. But when we think about Herceptin biosimilar erosion, I think we have seen things like Phesgo and subcut Herceptin actually do provide some slowing of biosimilar erosion just because, obviously, those formulations aren't available with the biosimilar. And again, for some systems, they see that reduction in the amount of share time as a really big cost savings for their overall system, and it's not just about the drug.

Yes. Thanks a lot. Emmanuel Papadakis would be the next one from Deutsche Bank.

Emmanuel from Deutsche Bank. Perhaps I could take a couple of big picture ones if I may. Margins, you're back up to mid-20s on R&D spend as a percentage of sales. You're now at about the right level or is that going to continue trending up as you shift the mix from other spend lines to R&D? And overall, thinking about the outlook for pharma margins over the next few years, you're back to growth next year. You have been slightly improving over recent years. Is there any reason we should not expect that to continue and even accelerate? You mentioned wanting to keep investors happy. Could we go through the mid-40s over the next few years? Is there any reason it couldn't get to the high 40s over the mid- to long term? Help us think about the natural ceiling. And then we know you're pretty plugged into Washington. Put your political hat on, how worried are you about something in form of price negotiation hitting Part B anytime soon? I'd love your latest thoughts. And maybe if I could squeeze in one on the outlook in Tecentriq. You obviously have a negative news around the TNBC labeling. Is there any risk we see something similar imminently in bladder, given Keytruda was somewhat surprisingly converted to a full approval despite Merck's KEYNOTE-361? And does that mean we're actually looking at a few down quarters ahead of the uptake in adjuvant? And how fast could that uptake in adjuvant be next year? Is that something you'd expect to be relatively rapid? And likewise for the additional adjuvant studies you are reading out next year, commercial implications of those?

A similar question, Bill, was from Roland, Goldman, on the Biden administration proposal. If you could kindly handle this one.

Let's see. So margins -- let me try to get this right. First off, I would say we had, as I mentioned, a very large increase in R&D spend in the first half of this year at a rate faster than you should expect to see in future quarters or future years based on -- partly on some onetime effects on COVID therapies and also kind of some onetime step-ups. As you saw our pipeline has grown rapidly -- very rapidly, our late-stage pipeline. And so that's been a driver as well. But that won't recur at that rate. We do believe that though the R&D spend, we will continue to increase the relative R&D spend over the coming years. We think that is part of the -- achieving our ambition. We think that's more of what the world needs is -- and frankly, there are many opportunities to spend less on the outside in terms of things like sales and marketing because of changes in decision-makers in the markets where we sell our products. So I think that some of those trends are going to continue. In terms of overall pharma margins, I think what we've been saying for several years is that we expect our margins to be stable. I'm not -- I'll give you my personal opinion, I don't think raising margins for an innovative company like ours over time is really the right call because R&D innovation, it's our lifeblood. And today's higher margin is tomorrow's kind of pipeline gap. And so I think we're going to be conscious of that. As you know, we have bold ambitions not only in medicines, but also in diagnostics and in insights. And some of those fields like the insights field is a rapidly emerging field. And it's not a profit-making field at the moment. And I don't think that's true for Roche, but it's also true for the other companies that are in that area. So based on that combination of things, I think you should expect from us stable margins. But personally, I don't think I see our margins increasing while we're pursuing the strategy. And in terms of pricing -- well, the situation in Washington, obviously, that's been a hot topic for the last week. Maybe I don't know, Teresa, you've been looking into it more closely, and you want to give your view.

So the White House Drug Plan as it came out is sort of largely what we expected. And if you look at what the White House has actually proposed, it's more sort of a collection of ideas, so sort of a laundry list, if you will, of things that could conceivably be looked at. We -- they do seem to be focused more on really tangible demonstration products -- projects than truly sort of enacting sort of big huge change. We've also heard Biden's administration say pretty repeatedly that their goal would be to actually push something through via legislative change versus just administrative action. And that they do want to try and do that through the congressional budget reconciliation package. We are expecting that Congress will be pushing something through in the next few days that will probably largely be based on H.R. 3. However, there's likely to be changes to that through the reconciliation process as we -- as it moves through. Our goals in this process haven't really changed. Our goal is to lower out-of-pocket patient costs and to find more sustainable ways for patients to receive their medicines over time. And we have been and will continue to be in close contact with the administration and on both sides of the aisle to try and find -- to find ways to make that happen. Overall, we believe that the U.S. is going to continue to be a great source of innovation and be a great partner in helping to drive innovation, and price is only one part of that. But it's certainly something that we've been watching closely. Bill, I don't know if there's more you'd add.

No. I think our long-term view is that patients need relief. The out-of-pocket costs are too high in the U.S. And we stand at the ready to engage in constructive conversations about how to lower patient out-of-pocket costs and how to reform many of the areas, whether it's in Part B or Part D that could use reform. And I think we remain optimistic that in the -- when the senators and representatives actually get down to drafting legislation, there'll be enough calm minds that appreciate the value of innovation that will end up with something we can live with. And it may cost us something. In fact, it almost certainly will, but that's okay. We can live with that. And as I said, I think we have a strategy that's built for all seasons, and so we're confident in our plans in any case. And I think, let's say, Charlie, do you want to comment on the question about the bladder indication and concerns you have about that?

Absolutely. I think the question started with the indication in triple-negative breast cancer. And just to clarify, as you know, we voluntarily withdrew that indication in the U.S. with the FDA. To be clear, that was a procedural question around the accelerated approval process. It was not a question of the efficacy or safety of the drug in triple-negative breast because, in fact, IMpassion130 we saw a clinically meaningful 7.5-month improvement in overall survival and found that the addition of Tecentriq to nab-paclitaxel was safe, but it was around a procedural question. To that end, the drug obviously is approved in roughly 90 countries around the globe outside of the U.S., of which 86 are full approvals. Moreover, we're continuing to examine and further confirm the benefit of Tecentriq in triple-negative rest and IMpassion132, which is an additional study in the metastatic setting as well as in the early setting with NSABP E59. With respect to the bladder indication for first line, as you know, the ODAC voted 10:1 in favor of maintaining the accelerated approval of Tecentriq in first-line bladder. The results of the INVIGOR-130, our Phase III study, met its co-primary endpoint of an improvement in progression-free survival. And we anticipate the mature overall survival results next year in 2022, and we look forward to sharing those with you.

Thanks. Thanks a lot.

I'll take the question on adjuvant. Okay. Great. So thank you, Charlie, for that. And I think the question on adjuvant and how quickly do we believe that uptake will be. I mean we do believe that [ ONO ] will become the standard of care for patients with PD-L1 positive adjuvant non-small cell lung cancer. Because the adjuvant treatment setting is still developing, it may be a little bit slower on the uptake. But we do believe again, that this is practice-changing information and practice-changing data. And that as screening protocols get better, as diagnostic testing gets better, we will continue to see that opportunity grow.

Yes. Thanks a lot. I have to say so far, we did really well because we had basically 2 questions per person to ask questions, and we try to keep our answers short. And if we can just continue that way, that will be perfect because otherwise we have a long list of people who have questions so that we can really address everything. Luisa, you would be the next one thank you for your question, Emmanuel. Luisa, I open your line.

So on giredestrant, I wanted to check a couple of things. For the second, third line study, do you need to show superiority to gain approval? And in the adjuvant study, what duration of therapy have you chosen and why? And then maybe to come back to Bill on essentially that margin question. So I mean I think what we can see today, the increased diversity of the pipeline entry into new therapeutic areas is very striking, as is the high number of NMEs. But just now you're indicating stable margin within pharma and a lower rate of R&D acceleration than we've seen this year. So that really leaves the marketing spend. So I'm just trying to, I guess, square the circle. So should we expect an acceleration in marketing spend with a bit of a lag on that R&D increase but you will be very mindful about how you deploy those costs, and this is about entry into those new areas? So just trying to think about that growth in the marketing spend.

Thanks, Luisa. Maybe I can answer the last question first, and then I'll turn it back to Charlie or Levi for the questions about giredestrant. And yes, so in terms of the margins, so Luisa, yes, let me just clarify, when I was talking about the R&D. So the R&D spend through the first half, I think it was a 19% increase. And so I'm saying we're not going to continue to increase R&D spend by 19%. That was based on some onetime factors. But going forward, we do expect to increase R&D faster than sales and faster than sales and marketing and G&A and operations. So in other words, we're looking to really hold the line on M&D, G&A, operations, while we're increasing the investment in R&D. I mean that's basically our medium-term outlook. So I hope that clarifies it. And then let's see who's -- Charlie?

Charlie?

So I can start with regard to your question about the second, third line trial. I think Levi spoke to it well, namely the mechanism of action, the apparent superiority preclinically giredestrant compared to the other available drugs as well as, I think, the emerging clinical data. We're bullish on the potential in that second, third line study realizing, I think, the caveats that Levi mentioned earlier, as you get into later lines of therapy, the dependency on is receptor have to be looked at. But we're -- we obviously are looking forward to sharing that results and hopefully showing a superiority of giredestrant in both benefit as well as tolerability -- efficacy and tolerability compared to fulvestrant. With regard to your question of duration of therapy in our adjuvant study, it's 5 years as is typically the case in the adjuvant setting. And I think the tolerability of giredestrant, which really does appear to be superior, I think, will allow us to help patients stay on these therapies longer and complete the 5-year duration.

Yes. Thanks a lot. Thanks for your questions, Luisa. Michael Leuchten would be the next one. Michael from UBS.

Two questions, please. One for Bill. Just going back to the very beginning of your presentation, Bill, where you talked about the aim to increase patient benefit at half the societal cost. Just wondering how you measure that. Is that as simple as bringing down the cost per therapy? Or is that a separate group that replicates what maybe ISO does in the U.S.? Any color would be helpful. And then a question on crovalimab. Just wondering if you could talk about your views on proximal complement inhibition versus downstream inhibition when you think about crovalimab in the development program.

Great. Sure. So in terms of the vision and delivering much more patient benefit at half the cost of society, so we definitely are looking at the price of medicines but also many other factors that determine cost to society. So for example, if you have a therapy that cures someone of cancer versus a therapy that extends their life or extends progression-free survival by 6 months and then they go on to another round of therapy and then another round of therapy, you have an opportunity not only to provide a great benefit for the patient, but also much lower cost for society. So we're definitely looking at that combination of things. I think we take drug pricing very seriously. We've demonstrated that with the pricing of OCREVUS, which beat the standard of care, yet we priced it lower than the standard of care. We demonstrated it with Hemlibra, launching inhibitors, beat the standard of care. We priced it much lower than the standard of care. We've done this. We just think that in order to be a sustainable life science company, we have to build trust with the public, with the payers, and there needs to be a fair deal, and it needs to be perceived as a fair deal. And so that's our look on it. Let's see.

Maybe, Bill, before you -- before we hand it over to Levi, there is a question in the chat which goes a bit in the same direction here about your vision, where you want to see the company going forward over the next 5 to 10 years. And this is from Ben Yeoh. He was asking about any metrics or any thoughts on speed, empowerment, decision-making in the company, how it evolves and how many -- what the [ imprint ] is. Could you give us any kind of anecdotes of talent attraction and so on? So what is your vision on the culture of the company?

Yes. Yes. Thanks for the question. No, it's sort of simple. It's not easy to do, but I think it's simple to describe. I mean really, we have 2 big goals. I mean one is we want to deliver on our vision, our ambition for the world, which is that we can deliver many more breakthroughs for patients, much greater patient benefit at half the cost of society. There's many ways to measure that. I mean we're looking at that. But one of the principles we have in transformation is we're not going to allow sort of academic discussions about measurement to prevent us from pressing ahead with what we know is the right answer. And so that's delivering that vision is 1 of the 2 goals. And the other goal is that we are the best place for people who want to spend their lives investing in life science, the best place to work and not the best place not only because of the extrinsic things that people think about, but especially the intrinsic things, the ability to do something great. And we believe that everybody has that right. If they're going to spend their life at a company, that we give them an opportunity to do something great to do things that they can be proud to tell their friends and family about, that Sunday nights that they're excited to come to work on Monday, that they don't have that dread of like, "Oh, no, another week." But they're -- on Sunday nights, they're like, "Man, I'm ready to get back. Let's go." So that's the culture we're trying to create. I welcome input from anyone else on the panel. I mean, we see the evidences, I mentioned some of those in the presentation. We see it from small teams, from large teams. People are excited about it. One thing I will say is that when I shared some of those figures on productivity, you saw lower head count, right? But we've not had -- we've not engaged in a cost-cutting exercise that I'm aware of in the last 5 years. I mean, we've been basically going around and saying, how do we make people's jobs better? How do we take a situation where we have, let's say, 3 people with 3 different roles serving a customer when actually 1 person could do that? And actually, it's better for the customer, and it's better for the person. They're not bumping into each other. They're not saying, "Oh, I'd like to help you with that, but that's actually somebody else's job." They say, "Yes. No, I can do that." We're creating human-sized jobs, which means big jobs that carry a lot of responsibility and weight. And people are excited about it. It doesn't mean that the change is always easy or that every day is a happy day, but I think we're making good progress. And I think we have a long way to go yet. So very excited about it.

Yes. Thank you. Crovalimab, Levi or...

Yes. I'll be very brief. It's a great question about proximal versus distal complement inhibition. And so I'll just say yes, it's an important question. Actually, the specific implications for whether it's crovalimab or any other medicine depend on the indication that we're pursuing. And there are some indications, which -- and Paulo could, for example, add in the ophthalmology sort of immune side, where we have a particular opinion on that. In other cases, we actually -- we know that it could be important, but we just need to gather data. So we are mindful of the specific elements of complement that may be gating in certain diseases versus others, but it will depend on the indication. I don't know, Paulo, if you want to give a specific ophtho example just for illustration.

Yes. Thanks, Levi. And I agree with that. And just to backtrack a little bit, usually, people look at the complement system, and it's been a little bit ignored in immunology. But it's actually one of the most complex biological systems in our body. It ticks over at a constant rate every day. It has over 20 proteins in it, multiple cascades. So to think upstream, downstream maybe is a little bit simplistic. What is clear is that both from genetics as well as from, yes, demonstration of clinical benefit with specific molecules to either targeting C5 or C3 or C1q, certain conditions are dependent on that particular protein more than others. So in ophthalmology, for example, it's been known for years, one of the biggest genetic risk factors for geographic atrophy is actually our molecules connected to the complement system. And as you may recall, we had our own program targeting component factor D, which did not work. And now just a couple of days ago, there were top line data showing that actually, molecule targeting C3 appears to reduce the size of lesion. So it is a little bit more complex just thinking upstream or downstream. I think for each condition, we have to go 1 or 2 layers deep to really find out what is the best molecule.

Thank you. Next one would be, thanks for the question, Michael, Andrew Baum from Citi.

A couple of questions. First on business development, Roche has a very healthy balance sheet. You've historically pursued bolt-on acquisitions of a small size. I'm assuming that's not going to change. But I was hoping you could comment on the outlook from the FTC, which is becoming a lot more muscular and there's some legislative risk ahead. So from the perspective of the industry, how concerned are you that this could severely impede business development going forward, thinking about your own recent experience with Spark? Second for Levi, one modality we're obviously interested [indiscernible] exploded is mRNA, I'm thinking less from vaccines for you but for cytokine delivery or exploring new therapy. I'm curious whether you believe you have the technologies in-house or whether this is something that you're interested in building out. And then finally, a question on gene therapy. There's obviously been a number of deaths associated with competing products. The FDA had a recent panel meeting to discuss and so on. Could you talk to how you see the regulatory hurdles for gene therapy in light of some of the recent news flow?

Over to you, Bill.

Great. Thanks, Andrew. Yes. So the question about business development, I think as I mentioned when I was talking about some of the partnerships that we've signed, our preference is to go early. I mean, that's why the vast majority -- I mean, numerically, it must be 97% or something of the deals we do, partnerships and acquisitions, are early stage. And I think -- so we're -- your point about the FTC, it's probably not going to have a big effect on us because so far, they haven't seen or seemed to intervene in really early-stage things that are high risk. It tends to be more where there's a known product perhaps with data or obviously something bigger than that. So I'm not sure that's really going to affect us very much. It may more affect the strategy of other companies. I'll turn it over to Levi on the questions about mRNA and gene therapy.

Great. So thanks, Andrew, for the question. Several of us, of course, mentioned the breadth of modalities that we have -- actually, that was primarily focused on molecules, assets that have reached clinical stage development. What we didn't talk about at all is the breadth of modalities that we're using -- that we're looking at in our preclinical setting. And so there, we're looking at even a broader range of potential platforms, including -- actually, one possibility is mRNA-based ways of generating asset. And of course, the converse is ways of inhibiting mRNA as opposed to protein. So there's a lot that we're looking at. I think the guiding principle here is that, as Bill mentioned, it really starts with the science, the disease biology. What do we understand about the disease? What does it need therapeutically? And whatever it needs, we're committed to figuring out how to do it. And if it's something that we already have in our wheelhouse, that's great. But if it's something that we need, as we showed even recently with our deal, for example, with Adaptimmune, then we'll partner to bring it into our wheelhouse. But we'll certainly -- but we start with the disease biology, understanding of the science and then we figure out what the modality solution needs to be. And maybe I'll let Paulo comment on the gene therapy question.

Yes. Thanks, Levi. You're -- this is actually a particularly sad day. There was a fourth patient that died on the Audentes trial just today. It was just announced. I think it just illustrates how young of a technology this still is. In lots of ways, we're learning how to manage gene therapies almost like on a disease-by-disease basis, and there's lots of factors that I think play into that. And we're, again, learning together with the FDA on how to manage those. Part of this has to do with the acute immune response to a viral vector, and different viral vectors seem to have different safety profiles. And obviously, again, we probably need new and better viral vectors. Some of this has to do with -- clearly with dose, how frequently and how high we're dosing. The route as well, direct administration to the CNS appears so far not to be a very good way to do things. Vectors that target particular organs like the liver or the kidney or the heart. Again, I think we're learning how to manage this. If you recall, like 20, 30 years ago, this is actually what brought down the first generation of gene therapies. Now of course, we're in a different place right now. We're much more sophisticated. And fortunately, a few of these gene therapies have made it all the way to the market, and they're having an impact. But it's still a cutting-edge technology. And therefore, the need for being really, really keen on monitoring safety and making sure we're doing the right things by these patients is going to be required in the short term and probably in the medium to long term as well.

Yes. Thanks a lot. So we have 21 minutes. We have so far 6 more people on the telephone. And I hope, Andrew, could address your questions. Richard Vosser would be the next one from JPMorgan.

A couple of questions on the gantenerumab data. So on the open-label extension, there were 38 patients that sort of didn't make it's out to the 36-month data point. And I think since the open label started, there was sufficient time for them to reach this time point. So sort of what happened to them? Was there any reason for discontinuation? Or was it that they just didn't get there? Secondly, on cognition for those patients, so those patients that achieved AB negativity over 36 months, how did that cognition benefit change relative to those who are AB negative at the start of the trial in the open-label extension, if I'm reading that data, right? And then second question, just very quickly, just on glofit in DLBCL. You mentioned going upfront with mosun on top of Polivy. What's the -- I understand you could use glofit for more severe patients. But what's really the point of using it later on if mosun is going to be better early? So why continue with glofit?

Yes. Paulo, do you want to take that?

Yes. Really, really great questions and perhaps just to take a step back because most might not know the history behind these open-label trials. So basically, back in 2015 when we declared futility for SCarlet RoAD, this was a very conscious decision to allow patients who wanted to continue to be dosed with gantenerumab and wanted to contribute to science to be on this trial, right? And over the years, we've opened it up for other trials as well with Marguerite RoAD, et cetera, and we upped the dose, as I mentioned. But it's really on a voluntary basis because these patients essentially know that this trial is generating biomarker data and safety data. But as you'd expect for a trial or for trials that now have been running for, some of them, 7 years, you will see some people dropping out. That's just the nature of these things. And it's kind of random, the reasons why. Some are for personal reasons. Some are because they want to get into other trials. But because it was an open-label trial, we didn't put any specific measures in place to ensure that people have to stay on this study. Of course, we all -- we collected as much data as possible, including safety data, to make sure that, yes, the things were going well with them. So that leads to the answer to your second question, which is regarding the efficacy data. Because of the way the trials were designed, this is not a controlled experiment. And because of the dropouts, et cetera, all the readouts we can put into these open-label trials are exploratory by nature and very tentative, right? We've presented and published that when you look at the efficacy data of cognitive outcomes in the open-label trials, there are trends that favor gantenerumab, right? But again, this is a very uncontrolled heterogeneous population. Some and people have been on drug for 7 years, and people have been on drug for 4 years. I wouldn't read much into it, frankly, and we didn't look particularly at the patients that were amyloid negative, again, because we're talking about a very, very uncontrolled small data set. So I hope that answers your question.

Charlie, do you want to take the future strategy questions on DLBCL first-line combination?

Absolutely. No, that's a great question about the bispecifics. And let me be clear, we're committed to the utility of both mosun and glofit. I mean, obviously, as you're aware, they have similar modes of action. But as Levi described well, they have very different structures and then, as a result, different profiles. And what I think, as Bill elaborated, I think they allow us to tailor therapy to the disease, the patients, the nature of where the care is delivered and the unique patient needs. Obviously, mosun is showing high and durable response rates and fortunately can be delivered safely in an outpatient setting. And we're anticipating filing mosun in -- initially in relapsed/refractory follicular lymphoma and working on it in combination with Polivy in second-line large cell lymphoma. Glofit, in contrast, really has -- what we think has the potential to offer best-in-class efficacy, which at least the data to date suggests that it may be comparable to CAR-T but with obviously easier applicability as compared to CAR-T. And so we think it's going to be an important therapy for diffuse large cell lymphoma. We're planning on a filing in 2022 in relapsed/refractory, but we are also looking at combinations with glofit, including now building on POLARIX in first line. So really, we think these 2 molecules will each add to the arsenal of therapies we now have for large cell lymphoma, both important and both, I think, have unique opportunities to expand the outcome for -- improve the outcome for patients.

Yes. Thank you. Richard Parkes, Exane, would be the next one.

Hopefully, you can hear me okay. A couple on Alzheimer's and then one on business development. So I wondered if you could update us of the feedback from the FDA of the potential to bring forward gantenerumab filing time line. I think -- I believe you were planning a meeting with the FDA to gather feedback on a specific strategy, so wondered if that meeting had occurred. And if you could give any insight feedback on feedback from FDA there, that would be helpful. Secondly, just a follow-up on the tau programs and semorinemab. It feels like you're encouraged by what you've seen in that Phase II program, but maybe need to see an improvement in order to progress to Phase III. So I'm wondering if we should think more about the second tau program you have targeting the mid-domain. And maybe you could outline for us why that might be differentiated based on that binding. Finally, just on business development. I think you kind of flagged that you've been eschewing sort of larger bolt-on deals in favor of smaller licensing-type transactions. But when you look at the field of what the opportunities out there at the moment, do you think you can maintain the very high level of activity you saw in 2020? Or should we be thinking about alternative ways for you to return excess capital to investors?

Maybe I'll answer the last question first, and then we'll turn the Paulo for the questions on Alzheimer's. Yes. So in 2020, yes, we did a number of late-stage deals. And that was -- I would say, with respect to timing, it was opportunistic. With respect to strategy, they were all on strategy. And in 2021, we just haven't seen that same picture emerge. But again, I would say 2020 has been more the -- or was more the exception and 2021, not so different. In 2019, we had many fewer deals. So I think that was more serendipity in 2020. And with respect to the returning capital, I think I've already commented on that. We're looking to maintain our margins stable, increase investment in R&D and be more efficient, more effective everywhere, including R&D. Paulo, do you want to...

Yes. Thanks, Bill. So to your first question around gantenerumab, I appreciate the question and the curiosity, but I have to be pretty clear. I mean, we're not going to comment -- I'm not going to comment on any ongoing negotiations with health authorities for any reason. Obviously, these are private conversations for both their purposes and ours. And let me just say this, I think it's encouraging and good that the FDA is showing openness to looking at drugs in Alzheimer's disease and trying to find ways to get them to patients faster. I think it's -- patients do need more options than they have now even with the recent approval. They need more drugs that target amyloids, with different administration routes, et cetera. So yes, I think it would be great if that was possible. But obviously, at the end of the day, we want to deliver medicines to patients that have a comprehensive data set, with a robust safety and efficacy profile that people can make informed decisions about their treatment. That is really what our commitment is, and that's why we're waiting for the readout of the gantenerumab trials late next year. Anything else is a matter right now for speculation, I'd say. So regarding tau, I mean, I am -- let me say, I am excited because when we think about the scale of this challenge, which is to get a monoclonal antibody into the CNS to target a misfolded protein traveling in between neurons, it's pretty remarkable that you see benefits happening. And that's why it's exciting because that in and of itself, the scientific challenge is huge. Secondly, the need for more drugs in Alzheimer's hasn't gone down. I think it will continue to be there, and tau is a hallmark pathology. So I think that's exciting as well. Of course, there is a number of question marks, right? Why didn't this work in earlier lines of treatment or in earlier patients? Why haven't we seen really a functional benefit to go along with the robust cognitive benefit we're seeing in moderate patients? Those are all really great questions, and those are the ones that we're going to tackle in the months ahead to try and find out whether this signal is real and what does it actually mean for these patients. Your final point was about the 2 tau programs. Yes, so we do believe in the same way that for amyloid, it took a while for us to find out which epitopes and which particular antibodies would be the most potent and most efficacious. I think for tau, it's a little bit the same story, right? So I think it's great that we have 2 differentiated assets in our pipeline, one targets the N domain, one the midterm domain. Both target all phosphorylated species of tau. Whether that makes it different or it doesn't, the data will tell. But our commitment is that this is such an important target that's so relevant for Alzheimer's and for other neurodegenerative conditions that we want to make sure that we don't miss out on really testing this hypothesis for the field.

Yes. Thank you. Stephen Scala would be the next from Cowen.

Can you hear me, Karl?

Yes, clearly. Thank you.

Two questions. The slide deck says that faricimab and Lucentis PDS approvals and rollouts would occur in 2022. Bill, you stated that faricimab and Lucentis PDS should be approved by this meeting next year. I believe in the second quarter slide deck, a chart said approval was expected in 2021. So it seems that there has been some sort of delay and perhaps a lengthy one. So I'm wondering if you could elaborate on this dynamic. And then secondly, on giredestrant, did Roche learn anything from the PARSIFAL trial showing fulvestrant was no better than letrozole when combined with CDK4/6 in first-line breast cancer?

Yes.

Steve, thanks for the question. Yes, I think we were just summarizing things on the ophthalmology program. There's no delay on PDS. PDS will be almost certainly approved in 2021. And yes, so I think we were just saying that both of them will be approved by this time next year. But we -- and I don't know, Nilesh, do you want to just reiterate what we've provided in terms of timing? Because I know it hasn't changed.

No. It hasn't. So for faricimab, we've provided quarter 1 guidance to 2022 for the first approval. Then obviously, through Europe, we'll be getting staggered approvals. For Port Delivery System, we have said Q4. And actually, I think we've communicated October, end of October as a PDUFA date for FDA and then later into 2022 for Europe.

Yes. Thank you. On the third, Charlie? Sorry, Levi.

Maybe I'll start on it. Yes. So thanks for the question. So the straight answer is, of course, yes, we learn from every study. But specifically, there have been a couple of informative studies in this space. One is the one that was mentioned, suggesting maybe a little difference in combination with the CDK inhibitor. There's also been, for example, the FALCON study, which has suggested differences in various modes of ER inhibition. And I do think that we -- coming back to one of the reasons why we're interested in giredestrant is there's an opportunity to ask the question in a different way because of the mechanism of action. So not only are we causing degradation of the SERD, but because of a distinct effect biologically, we're suppressing the transcriptional activity of the estrogen receptor. And this is an important -- it's subtle, but it's an important distinction because we know that in -- particularly in the early disease settings, but maybe in metastatic disease as well, it's not just the ligand-dependent actions of the estrogen receptor that are important for the oncogenesis. It is also ligand-independent actions of the estrogen receptor. So a medicine like giredestrant can go after both of those mechanisms and to suppress them both. So we've learned from -- we've learned and we have cautionary tales from the existing data, but we also think that the biology that we're able to go after with giredestrant is separable from some of those earlier studies and, hence, provide a rationale for the program that we're pursuing.

Yes. Thank you. We have 3 more questions on the line, and I have also 1 on the anti-infective part. So next one would be Keyur Parekh.

Hopefully, you can hear me okay.

Yes. Perfect. Thank you.

Great. A couple of questions, please, the first one for Bill. Bill, I think in response to Jo Walton's question about R&D, you seemed to suggest confidence in R&D productivity improving over the course of the next few years. You are factually stating that your R&D spend is going to go up. Should we, therefore, feel confident that your revenue growth over the next few years will be higher than what you've delivered over the last decade, just multiplying -- putting the 2 things together? Then second one kind of for -- kind of on the ophthalmologic franchise, there seems to be a degree of confidence on both faricimab and kind of Lucentis Port Delivery System from a commercial perspective that is not being reflected in kind of consensus numbers. So I'm wondering what gives you the confidence on what it is that we may be missing kind of as a group of analysts there? And then just lastly, on the SERD kind of adjuvant program. A couple of your peers have gone into the adjuvant setting with a much broader kind of program than the one you are kind of specifically alluding to today. Given your confidence on the safety, tolerability and the potential for greater efficacy there, I'm just wondering why the adjuvant program isn't a lot deeper and broader.

Yes. Thank you. Maybe you can, maybe this time, first start with the SERD because actually, I think it's the other way around. But please correct me, Levi and Charles.

Well, that's exactly right. So the trial that Sanofi has launched is actually limited to patients who were essentially intolerant of aromatase inhibitors. In contrast, because of our conviction and belief in the efficacy and safety of giredestrant, our trial actually is in the broader population of patients who need upfront adjuvant therapy. So really, it's the opposite. The other trial is limited. And obviously, one can only speculate on the rationale for that limitation.

Yes. Thank you. Nilesh, you would like to take the faricimab and that you definitely will do better than what the market expects from you.

Yes. Well, that's what certainly our hope. And thank you, Keyur, for the question. I think we're very confident that these are true innovations compared to current standard of care. So on the Port Delivery System, you're seeing a twice-yearly injection provide the same type of outcomes as monthly injections of Lucentis 0.5, which have never been seen before. And I think that's going to be something important for patients, for their caregivers and also for society so people won't have to come back. And on faricimab, we're seeing for the first time a distinct pathway, not a different target within the same pathway, but 2 different pathways are really delivering what we've seen from preclinical science all the way through now to clinical science that vascular stability is driving some of this durability. So we're very confident that these can become new standards of care. I do believe that, obviously, a lot of people are concerned about biosimilar impact in this area. But I think as Karl said, in this area, we've had a -- certainly, in the U.S., 50% of the volume of the market has been in an Avastin setting anyway. So maybe there's some different ways to look at this marketplace.

Yes. Maybe the combination of a recent launch that didn't go so well in ophtha and just the biosimilar question and people are being cautious. But you all are the analysts, and we just make medicines. So you'll have to tell us what the difference is. Your question about if we're going to invest much more in R&D and we're going to have higher R&D productivity, then doesn't that mean we're going to have higher revenue? I think -- let's say, our hope, the reason we get up in the morning is we want to rewrite the medical textbooks. We want to have the biggest payoff for patients ever, and we're really proud of our history. Roche is celebrating 125 years this month, the proud history of Roche, the proud history of Genentech delivering breakthroughs in biologics and cloning cells and converting that into something that helps people. And I think for us, I speak for all of us, yes, we want to get some new medicines in the history books. And when we do that, I'm sure that will be good for shareholders, too. But we're not counting our revenues right now. We're focused on delivering those molecules, and we'll let the revenues follow.

Yes. Thank you. Thanks for your question, Keyur. Peter Welford, [Operator Instructions] from Jefferies.

Just a couple. So firstly, just on immunology. Curious that -- Roche obviously works in a lot of areas, including in some areas of immunology. But there was little, I guess, relevant -- or discussion around either etrolizumab in Crohn's and also perhaps any other efforts you have ongoing in any other of the -- or the sort of unmet medical needs still, thinking in terms of atopic dermatitis, et cetera. I guess curious on Roche's view in that area and whether that is an area where you think Roche is currently underweight or an area of deliberate focus or rather not to be focused in. Secondly, just coming back to giredestrant and the pivotal Phase II readout that's coming soon. You presented at ASCO the VERONICA study, where I think you saw it was a failed trial. But I think you saw with fulvestrant and plus or minus Venclexta roughly a sort of 2- or 3-month PFS. Do you think that's sort of representative of what we should be thinking about in the pivotal Phase II for giredestrant, for fulvestrant? And I guess curious what you think perhaps then giredestrant could do as the active arm to demonstrate the superiority you're looking for in that profile to potentially take it to market based on those Phase II data.

Paulo?

Thanks, Peter, for the question. Maybe I'll take the immunology one first. So I totally agree with you. Immunology is a fundamental area of biology. And there have been many, many medical breakthroughs in the past for numerous conditions. So if anything, now the bar is that much higher to differentiate and create more medical value for patients, right? It is something we're very committed to. We have a long, proud history here. And you're right, we have a few programs which are ongoing right now in Phase III and pivotal trials. Obviously, we're waiting for the readouts of etrolizumab in Crohn's later this year. I think I mentioned we have our Gazyva program with lupus nephritis and in SLE. At the same time, we have our program with Pentraxin-2 in IPF. So I guess a common thread there, if you want, is that we're looking to really break that efficacy ceiling right now. So we're going to have to be a little bit, if you want, more selective about which mechanisms and which molecules we're going to take forward because right now, there are a lot of great medicines out there. So for us to really differentiate and do what Bill said, which is rewrite new pages in the medical textbooks, we're going to have to try different things. Fortunately, we have an amazing research pipeline, many new mechanisms, and we are committed to doing this. It may take a couple of years until you see new molecules coming in, but we're not lowering the bar, right? We're not just trying to develop the fifth anti-integrin or whatever else, anti-TNF. We want to do things that are different and fundamentally bring more value to patients.

Yes. Thank you. The third, Charlie?

Sure. And with regard to the question about giredestrant and namely our pivotal trial that will read out next year of the head-to-head comparison with fulvestrant in second and third line, I think, as you've heard, the evidence that our molecule is 7 to 15x more potent preclinically compared to other potential degraders or inhibitors of the S receptor, the fact that, as Levi pointed out, that our molecule by virtue of its mechanism action works on both ligand-dependent and ligand-independent mechanisms of activity. And the latter front, I think, is important, particularly in later lines, where we're more likely to see ESR1 mutations. So I think the data really suggests a good opportunity for giredestrant to prove superior to fulvestrant. We'll find out next year. But I think the data, the safety, and I would draw your attention to our presentation at ESMO this coming week, which I think will provide some additional clinical data of interest with regard to the efficacy of our molecule.

I just think -- because it hasn't been explicitly mentioned on fulvestrant with several questions, fulvestrant has historically been limited by bioavailability. You just can't get enough fulvestrant and to maximize the effects against the estrogen receptor. And we think one of the features of giredestrant is the activity and bioavailability they were able to get with once-daily dosing. So again, we have to wait for the data, but that's another reason why we are looking forward to the results.

Yes. Thank you. Sam, you have the privilege to have the last question in the queue. And then I have one more, which is over the chat. Sam Fazeli from Bloomberg, over to you. Thank you, Peter, for the questions.

Right. Am I unmuted? Can you hear me?

We can hear you, yes.

Excellent. I just -- I'm going to start with a very simple one, Actemra. Are you supply constrained at the moment still? And if so, is there any way that you can change that in terms of getting more product out in the market? Then on AT-527, you have MOONSONG that's got a time -- 22nd September for completion. And whereas the Phase III trial MORNINGSKY is 3rd of August, which seems like kind of in reverse in terms of the dose ranging study being -- coming later than the Phase III study. So if you could just explain a bit where we are with that and when you might actually expect. I know you said the end of the year. And then lastly, on the Brain Shuttle gantenerumab, lots of related questions here. Any risk of ADA, antibody -- antidrug antibodies? The high concentrations in the brain, have you seen any risk of increased inflammation more than gantenerumab in the few studies that you've done in -- few patients that you've done in Phase II/III? And what is the target product profile here? If gantenerumab works, are you aiming to reduce the dose? Or are you hoping that the Brain Shuttle will actually help you increase the efficacy? I know there -- obviously, it could be a bit of both. That's it.

Yes. Thank you. Actually, you took 2 of the last remaining questions out of the chat. So we can actually close with your questions. Yes, supply constraints of Actemra.

Yes. So we are supply constrained on Actemra. We've -- basically, since the early days of the pandemic, literally, even in March of last year, we have begun supplying Actemra for pandemic use on a sort of experimental basis to countries requesting it. We've -- I'm very proud to say that we've been able to supply well over 1 million doses. And hopefully, we've saved many people's lives with that. But the simple fact of the matter is, it's a monoclonal antibody. There's only so much production capacity for monoclonal antibodies, and we've also been producing the monoclonal antibody cocktail. And so basically, yes, we've had all our plants running at 100% capacity since the beginning of 2020. That continues to be so. As we saw the potential for demand to really outpace supply, we signed contract manufacturing agreements both with a major contract manufacturer, the largest in the world, as well as with Novartis, our friends down the river in Basel to make more Actemra. And actually, that Actemra is coming online beginning in October. So we have probably significant additional supply coming in, in October, which will help. We were doing fine on supply up until the Delta variant hit and just the magnitude of supply, we've supplied, I think -- through July, we had supplied 60% of our pandemic supplies went to low- and middle-income countries, including India. So we've really done our best to meet the need of the world but just coming up a little short. But we're hopeful that most patients that were ongoing Actemra patients were able to kind of make it through or maybe miss minimal number of doses, and we'll be able to get resupplied in many places in October. And let's see, the other questions were around faricimab and AT-527. Barry, you want to...

Thanks. I'll take that question. So I think the question was around the relative timing of the Phase II MOONSONG versus the Phase III MORNINGSKY. They do partially overlap each other. That message on the slide where we said that we would expect the outcome from MORNINGSKY towards the end of this year, that is correct. That's where that will happen. We intentionally started them with a partial overlap based on preliminary data that we had from in vitro models and from the prior HCV program knowing that if that proved to be correct, that would optimize our time to market and our time to patients. But the MOONSONG is there to make sure that we are picking the right dose because it is critical for us that we bring forward the correct dose so that the medicine has the best possible impact for patients. So at the moment, that's where we stand. They do partially overlap, and we're still aiming for data from MORNINGSKY at the end of this year.

Yes. So we are coming to an end of...

Yes. Sorry, there was a question on Brain Shuttle. Is it okay if I can take that? Yes. So that's a really interesting question, right? This is a technology that is still very much being experimented on and being developed. And this is our first human trial with this platform, right? So as with any new thing, the questions you asked about antidrug antibodies, et cetera, those we'll have to find out as we move along. Of course, we monitor safety of patients very, very carefully, and that's one of the things we're going to be monitoring. The other question, I think, you mentioned was around the potential for increased inflammation in the brain. Now really, nobody knows what causes ARIA. To this point, there is some relationship with dose. There's some relationship with stage of the disease, et cetera. Is there a potential? Yes, and it's something we're going to monitor again. And right now, just to remind everybody, we're just starting this Phase Ib trial with 120 patients in prodromal to mild Alzheimer's. So data from those trials -- that trial is going to be really informative. Your final question is a harder one. What do we expect from this molecule basically? So what's the TPP? And I think it might be a variety of things, and we will really let the data tell us which one of them is more likely. It could have to do very simply with dose, with the amount of drug, with the duration of dosing, with the frequency of dosing. It might be that we can achieve faster amyloid reduction, and that does translate into clinical benefit. Again, there's no evidence now, but that is a possibility. One interesting twist to this story is that actually, when we look at the brain distribution patterns of the normal gantenerumab and the Brain Shuttle version, the Brain Shuttle version goes to different places. It essentially floods the whole parenchyma almost evenly with drug, whereas the normal version goes mainly through the choroid plexuses. So it has more of a cortical to then deep brain gradient. So it might be that, that causes differences in efficacy. But again, it's very early days. We're very excited about the possibility because, again, I think there is concurring evidence that amyloid is an important target in Alzheimer's. And therefore, we just essentially are going to let the science tells us what to do next from this molecule.

Many, many thanks. Just from my side, thanks to Bruno Eschli, who took the lead on this event. Thank you, Bruno. I know that there have been some night shifts involved in making that. The same is true for Loren in the United States. I wanted to thank you for your time today because you also worked on all the setups. I wanted to thank the back office to make the whole thing running. I wanted to thank you in your -- for your interest in Roche, participating in the [indiscernible]. I've seen that many people really actively participated in that [indiscernible]. So that is really good to us. Bill, maybe 1 or 2 words from your side. Thanks for your interest in Roche. Wishing you all the best, and over to Bill.

And again, my thanks to everyone who's contributed today. And thanks for your questions and for your interest in Roche. As you can tell, we're really excited about what we're going to try to deliver for patients and for the world. And we can't do it without our investors and your confidence. And so again, many thanks. Wish you a great rest of your day, and look forward to seeing you all again at another event soon.

Thank you. Bye-bye.