[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.

Yes. Thanks a lot, Henrik, for the introduction. welcome to our Q2 call. I hope you are all safe and well. Actually, we have a new setup here, as we are all in one room, which is really exciting, I have to say. Hopefully, the times of Zoom are now behind us, but we have social distancing in that sense. So it's all checked, but it's really exciting to be here with the colleagues and to see how they're doing. If you have difficulties to ask questions via the Zoom setup, of course, you can also drop me an e-mail, karl.mahler@roche.com. And for the Q&A session later on, maybe we can limit the questions per person to 2. We also try to be short and crisp in our answers. So that we, let's say, maximize the time, later on, also for that part. Personally, I think we had really an excellent first half product performance, the new products was excellent. Diagnostics business underlying was excellent. And tomorrow, we will have some roadshows where we have -- where we can connect in the U.K., Switzerland, France, U.S. and so on. With this one, over to you, Severin.

Thank you, Karl. And also a warm welcome from my side. Happy to dive into a strong set of half year results. So let's get right into it. Here we go. Group sales, as you have seen, up by 8%, really strong growth by Diagnostics with over 50%. And what I should say here is that, yes, that was, of course, also driven by our COVID-19 testing. But as you will see later, really strong recovery of the business, which is good to see. Pharma, likewise, after a negative quarter in Q1, we have seen a recovery with positive growth in the second quarter. And importantly, the growth is driven by our newly launched medicines, up by 30%. So really, all in all, a result, which is, of course, heavily impacted by COVID-19. But the good news is that the underlying business is on a very good track. Core EPS up by 6%. And then let me just highlight the pipeline. I mean, we speak so much about COVID that we tend to not look at all the medicine, the other medicines, perhaps so much. So I think it's fantastic to see that we had 12 Phase III trials initiated. And now 18 new molecular entities in late-stage development. This is a record for Roche. We've always had a industry-leading pipeline. And I believe with this new record now we are even better positioned for longer-term growth of the enterprise. Now if we go to the next slide, I won't cover that in detail, but I mean, we are actually quite proud of our many contributions, both on the Diagnostics and on the Pharma side. We making the fight against COVID-19. Let me just highlight here on a brief, we got 2 days ago, full approval in Japan. We recorded already about CHF600 million sales in the first half. We expect a similar turnover for the second half. We see continued demand for Actemra. And as you know, we have a small molecule in development with our partner, Atea. On the Diagnostic side, I'd just like to highlight the base effects which we see versus the previous year. We've seen a very strong growth versus the first half of 2020. And this is really due -- also due to the fact that we didn't sell much yet in the first half last year because we were ramping up capacities. That changed in the second half of the year where we could provide much more testing to our customers. And as a consequence, what you will see year-on-year, you will see a decline of growth versus the previous year. Now in terms of absolute growth, what will we see for COVID-19 testing will, of course, very much depend on how the pandemic evolves. But at any rate, we will have a significant base effect when we look at year-over-year growth. Right. So if we go to the next slide, again, Pharmaceuticals with minus 3%, but already positive growth in the second quarter and 51% on the Diagnostics side. On this slide, you can see the quarterly development, and you see the strong recovery now in Q2. Part of it, of course, is a base effect because we were declining in Q2 last year. But you can see that beyond the base effect, the business is really getting back on track, and that's really good to see. Now here, the key elements for the growth. And this is really about the newly launched medicines on the Pharma side and the strong growth in Diagnostics, which, as expected, is partially offset by the entry of biosimilars. And you can see here a very significant impact of CHF2.8 billion, which we assumed to decline, as expected, I should say, in the second half of this year. Here, I'd just like to draw your attention on the orange line. This is the new products and the development of pharma, the newly launched medicines. And you can see, again, the strong recovery in the second half -- sorry, in the second quarter of the year, and we expect that to continue as we go into the second half of the year. On the profitability side, we grew operating profit by 4%. You can see a decline of the margins of 2 percentage points. That's primarily a product mix effect with the overproportional growth in Diagnostics and some specific effects in the first half. But overall, we are holding up well. You also see good development on the operating free cash flow. So with this, let me conclude for the outlook, again, great progress in the late-stage pipeline. Also a number of new launches on the diagnostics side underway. So I think we can look with confidence into the years ahead. And based on the good results in the first half, we also confirm the outlook for the full year with low to mid-single digit sales growth, EPS in line with sales. and our plan to further increase the dividend in Swiss francs. Thank you very much. And with this, I hand over to Bill.

Thanks, Severin, and good to have a chance to connect with everyone. We had a very dynamic first half, and we're really pleased to have the opportunity to share the results with you. So looking at it from a revenue standpoint, as Severin mentioned, minus 3% in the first half. But I think very importantly, we started with minus 9% in Q1. In Q2, we were actually plus 4%, and we think that pertains very well for the outlook for the quarters that we have ahead of us. Now you can see the biggest decline was in sales in the U.S., and that, of course, was driven by the biosimilar impact. From a P&L standpoint, it's a rather eventful P&L, in particular, I would draw your attention to the fact that we continued to save on cost of sales. So our volume was up 8% in the first half, but cost of sales was down 4%, G&A was down 2%. Very significantly, we dropped marketing and distribution expenses by 6%. We think for the full year, it might even be a little more than that. And we're taking that money and rerouting it into R&D. This is something that we've communicated already 2 years ago, our intention to do this, and we're doing basically what we said we would do. And we're using this money to invest in what Severin mentioned, the 18 molecules we have in Phase III or registration as well as substantial investments in gRED, pRED, Chugai and other parts of the R&D enterprise like Spark and FMI and Flatiron. We really believe this is our future of our company and that's what the world needs from pharmaceutical companies is more emphasis on innovation and less on the other things that are not the things that are directly driving patient benefit. So that's a commitment that we made and that we're keeping. And I think I'll have more evidence in the slides ahead of what that's going to mean for us in the future. This is the product view, and you can see very significantly at the top with Ronapreve leading the growth. Of course, 100% of Ronapreve sales were growth because there was no sales before the first half of this year. But you also see Ocrevus, Hemlibra, Tecentriq, Evrysdi leading up the list, and that's really good signs, and I'll show you more on that. And then, of course, at the bottom, you can see Herceptin, MabThera and Avastin. A couple of things I would point out, on Avastin you see a large green bar. So that was the Avastin declines in Europe. Essentially, it's gone away in Europe. And so the declines in the future should be much smaller. And then you see the large blue bars that reflect the U.S. We've now lost significantly more than half of all 3 products in the U.S. And again, future declines should be somewhat smaller. In terms of the overview of oncology, you can see that we still have some impact of biosimilars certainly in oncology and a small impact of COVID. But we think that we might be back to around 95% and 96% of normal in terms of oncology visits now in the major markets. That can be affected by Wave 4, Delta variant, et cetera. But on average, we think we're something like that. I would point out on this chart, you can see that Perjeta now is the largest oncology product that we have. and significantly Tecentriq past Avastin in Q2 to be the second largest. And so now we have Avastin Herceptin and Rituxan, a much smaller part of our total than before. Here's a little more detail on Tecentriq. And again, you can see a very encouraging growth trend here. We saw an impact last year from the pandemic, but now we've started to accelerate again. And this is really driven by the additional penetration of small cell lung cancer of HCC. You see Japan with important growth there, driven by the liver cancer indication. And the outlook is very positive as we have the adjuvant readout that we had in the first half, adjuvant lung, we'll be launching in the U.S., in Europe and Japan in the coming quarters, and we're really looking forward to the opportunity to bring the promise of cancer immunotherapy for the first time to patients with the earliest lung cancer. Hemophilia A and Hemlibra, another, I think, very encouraging growth chart. And as you can see, the growth rate today is really back on track with where we were before the pandemic. We had an impact, but it's slowly been picking up speed and Q2 was the biggest growth quarter we've had since the beginning of the pandemic. Again, we see further growth in every geography as patients continue to benefit from Hemlibra and tell their friends. And it's -- we don't really see an end of growth in Hemlibra, in sight. Moving on to immunology. Here, the story was partly about Actemra. So we had 12% growth in Actemra. As you know, we've now been recommended by the WHO worldwide. We received an EUA for COVID pneumonia in the U.S. as well. And we're working very hard to produce enough Actemra. I mean I will say that our margins on Actemra are relatively low in COVID pneumonia because a lot of our sales are to emerging market countries and our price is much lower in those countries where we're implementing. Yes, we've implemented what we call international differential pricing, which is based on an index of the GDP per capita for each country. And so our prices go much lower when you're in some of the emerging markets that are unfortunately experienced in these severe bouts of COVID right now. I would also just call to your attention at the bottom, to the fact that we started a first Phase III study in membranous nephropathy for Gazyva. So this is the second indication we're pursuing in the kidney space. We also have lupus nephritis study ongoing with Gazyva. And so those studies, we'll be reading out in coming quarters, and we're looking forward to hopefully bringing the promise of Gazyva to more patients with immune diseases. Moving on to the MS franchise. So this chart probably deserves a little explanation because, as you recall, in Q2 of 2020, due to the pandemic, we had a big reduction in infusions and these were many patients who were already on Ocrevus who deferred their infusions because they were concerned about the implications of taking a B-cell blocking therapy with a pandemic that led to then a rebound in Q3 and then the pattern continued because the dosing happens every 6 months. And so we saw basically a good quarter and then a soft quarter and then a good quarter and then a soft quarter. So when you look at Q2, what we see is actually quite encouraging, even though sales were slightly down from Q1, but we see that pattern being disrupted, and we look forward to strong continued growth for Ocrevus. Also, I will say, we see the level of switching new patients and switching is picking up. So in February, we believe it was about 75% of the pre-pandemic levels. And then in May, we think that number went up from 75% up to about 88%. So we're getting quite close to where we were before the pandemic, in terms of the level of switching and Ocrevus is maintaining 36% new-to-brand share in the latest data. We're really thrilled with that result. And I think it just underscores the importance of our world-leading data on disability progression, which is true both in relapsing MS and primary progressive MS. Also, I think very significantly, we're making great strides in accruing our high dose studies for Ocrevus. And I think for investors, this is really one to watch because the only thing that we can think of that would be better than Ocrevus in MS is a better Ocrevus, and that's really what the high dose studies have the promise of delivering. And this is based on the proof-of-concept data that we saw from the exposure levels in the Phase III program. So really glad that, that study has been -- or those studies have been very popular with investigators and patients and are crewing rather rapidly. Moving on to spinal muscular atrophy and Evrysdi. We're really encouraged by the uptake of Evrysdi by physicians and patients around the world. This is really in every country. And it's -- I think the benefit of a once-a-day oral formulation that delivers a systemic therapy. It's not confined to the CNS, but it delivers the benefit of the gene-splicing modification to every cell in the body. And we think that can be important for peripheral nervous system. I want to highlight the graph on the right. So this was the data that we presented recently at Cure SMA. And this is the first data that we've had in newborn babies. So these are presymptomatic babies treated proactively with Evrysdi based on genetic testing only. And what you see is essentially that after 12 months of treatment, 5 out of 5 infants, so 100% of these babies scored above 60; of 4 out of 5 scored 64, which is the maximum score. So that's as good as you can do; and then 1 of the 5 scored 63. So really, it just -- it couldn't be more encouraging. We'll have more babies and more long-term data in the months and quarters ahead, but this is a very strong signal. And I think it was very well received by the community. I want to move on now to ophthalmology. Again, we're very excited after many years of investment to be bringing not 1, but 2 new medicines for people with diseases of the eye, and in fact, we'll be launching these around the world in the next quarters. Starting with Q4, we believe we'll be launching the port delivery system in the U.S. And I think you've seen the data, so I won't go into that right now, but we really think these are game changers. They have the potential to deliver really the promise of this kind of therapy which has been there, that promise has been there for 15 years since the launch of Lucentis, but unfortunately, most of the time that promise isn't delivered because patients don't get enough doses. And so with faricimab, 75% of patients can have 4 doses a year and get the full benefit. And with PDS we see 100% of patients able to get the full benefit with 6 monthly dosing. And I just want to say that we got our first patient in a Phase III study with a 36-week dosing interval. So basically, it's -- that equates to about 1.3 doses a year, which is a really amazing step forward for patients. And again, we look forward to bringing that result in, in the coming quarters. Yes, I want to conclude with a couple of overview slides on the pipeline. Severin mentioned, we started 12 Phase III studies in the first half. It's been a very, very busy time. I think that helps to explain the R&D investment. But if you look at the list, there are some very important studies here, including Kadcyla and Tecentriq in 2 different studies in breast cancer. Also faricimab in RVO, fenebrutinib in relapsing MS. That's our BTK inhibitor. And then giredestrant, I want to point out, in early breast cancer. So this is adjuvant hormone positive breast cancer, ER/PR-positive breast cancer. This is really the largest breast cancer indication in the world, and we've been really excited by the data that we've seen with giredestrant. We have a late breaker at ESMO, you want to tune in for that, and we'll be sharing more data on giredestrant at ESMO. And then in terms of news flow ahead, if you look at the list on the right side of the slide, I think, again, some really encouraging things, 4 new NMEs with data reading out yet this year. and then very significantly Polivy, the POLARIX study, which is Polivy with R-CHP, which is a chemo-sparing regimen which is in DLBCL, and this will be the first major readout or at least hopefully, the first positive readout in about 17 years in DLBCL, and we hope to have that result yet in Q3. So that's a very exciting one and really will determine the future of Polivy. In 2022, it's an amazing list, starting with 3 studies of tiragolumab. That's our anti-TIGIT molecule. That's some really game-changing stuff for cancer immunotherapy, our best shot at improving on checkpoint inhibitors. And so those are really important. The first giredestrant study in second-line and third-line breast cancer and then gantenerumab in Alzheimer's disease, which I'm sure we'll be discussing further. I don't want to give short shrift to the many adjuvant studies lifted below the line on there, but we can talk about those in the Q&A if you'd like. And then finally, just today, we're very proud of our track record this year. We have a lot of green checks, including 5 positive pivotal studies, which is very encouraging. And I want to make sure I invite all of you to come to our Pharma Day on September 14. And with that, I'll pass it over to Thomas.

Thank you very much, Bill, and good morning, and good afternoon, everybody. I'm very happy to present the half year, Diagnostics' division performance. Now with sales of more than CHF9 billion, we had 51% growth in constant exchange rates. And the growth was driven by COVID-19 testing on the one hand, but not only, we had very strong growth in our base business or routine testing business, and I'll go into that on the next slide. And as you can see, all businesses contributed strongly to the very strong growth results. Now I would like to highlight one thing that is Diabetes Care, growing 10%. As I had mentioned already in our Q1 call, here we had a resolution of a dispute over a rebate in North America in Q1, and without this dispute, sales grew 6%. Now on this slide, you see, on the one hand, in the blue line, the Diagnostics division sales growth and in the orange line, you see the routine sales growth. And what you can see is that we are growing very strongly in Q2 with 31%. And this is partly due to a base effect that you can see in Q2 '20 versus Q2 '21, but not only, we had really strong growth in our business overall in Q1 and Q2 in our routine business. And I would say we really exceeded what our own expectations were in this area. Now we always said, even during the pandemic, we need to keep our eye on the ball when it comes to the rest of our business. This is the business that's going to stay around even after COVID. So seeing that growth, that's just fantastic because this will not go away, right? And so we'll continue very good momentum in that area. Now COVID testing sales were CHF2.5 billion. And as you can see, we could even beat Q1 with -- in Q2, CHF1.3 billion, even more than CHF1.3 billion sales. And this is due to the fact that laboratories are consolidating on the most automated platforms, and I think we are very well positioned there. And as expected and also on how we've reflected it in our plans in the beginning of the year and how we communicated it, we do expect demand for testing will decrease in the second half of the year. And this will depend very strongly on the progress of vaccinations, when people get vaccinated, and it will depend on the effectiveness of vaccines. due to the new variants. We see that with Delta, but there may be other variants that will come up and how effective are vaccines, that's the question. And depending on that, we'll see more or less COVID sales. But up to now, I have to say we are exactly on where we thought we would be around COVID testing for the first half of the year. The second element I would like to highlight is that, especially in Q3 and Q4, we had strong growth. So as Severin mentioned in the beginning, there will be also a base effect in terms of growth rates because we are comparing to a much higher base. Now looking at the regional sales. And here, you can see strong growth across all regions and particularly EMEA and Latin America. Looking into the different businesses. What we can see is that 34% growth in our core lab, which is amazing, strongly driven by immunodiagnostics and clinical chemistry. We still have very strong growth in the molecular lab. And if we look at it there, we see that even in Q2, most of Q2, we were still sold out on the 6800/8800 SARS-CoV-2 assay, where we saw a decline was in our more manual systems with MagNA Pure LightCycler. But as we have increased capacity, we are now in a situation that we can be moved away in June actually from allocation, and we can continue to supply the world as they need it. Point of care grew amazingly with 349%, point-of-care immunodiagnostics with more than 2,000%. This is really due to the rapid antigen test sales and also pathology grew nicely with 20%. Also a strong recovery in advanced staining business, but also companion diagnostics is growing with 15%. Diabetes Care already mentioned on the first slide. Now if we look into the P&L, we see that we more than doubled our core operating profit. We are growing at 137% and you see that cost of sales are exactly in line with sales growth. This was also our commitment when we said we're going to sell a COVID testing in line with pricing that we had prior to the pandemic. And you can see that reflected here that the volume growth is exactly in line with the sales growth. Now M&D is growing 8%. Here, M&D consists of, obviously, marketing sales, but also distribution. The growth is really coming from the distribution part. And the rapid antigen tests were actually quite costly when it comes to distribution in the countries because the volumes that we shipped were just so big. And this really hits the M&D line. And without these onetime effects, we would actually be flat in M&D. R&D is increasing because we're investing more into COVID products, but also other innovation. G&A, here, we also have a onetime effect that's significantly impacted. There were certain costs for admin functions that were not booked under admin in the past. And so we just clean that up and make sure that it's booked on the G&A and that's where this growth is coming from. Without that, we would also be more flat. So we're really ensuring good cost control across the different cost lines, underlying to make sure that if COVID becomes less, we're not stuck with certain costs afterwards. Now going to our Diagnostics SARS-CoV-2 portfolio. We have delivered 21 solutions to help fight this pandemic. And this includes molecular solutions, immunology solutions, digital solutions, both for central labs, but also in the patient testing settings. And this portfolio has really become a significant factor in supporting patients. And I can say we -- and our organization was really fast, but not only were they fast, they also delivered great quality. And this can be seen, for instance, in our antibody tests that are really performing extremely well and where we continuously see also more uptake. Now I'm going to go into 3 products on this list. Most of them you already know. And these are the green ones that we recently brought to the market. One is SARS-CoV-2 assay on the cobas Liat. So we had the duplex test or the triplex test on assay, before. Now this is a test just for SARS-CoV-2. And this is actually the first assay that has received emergency use approval for asymptomatic screening in the U.S. And obviously, this test can detect all variants. And just as we design the PCR assays and for other parts of the portfolio, we have certain safety measures built in so that they can recognize also newly -- new variants that may come. The second part is that we received now full proof CE mark approval for the rapid antigen nasal test, the self test. Now this is important because so far, all the testing that was done in Europe was only under special approval. And these special approvals were only limited to a very short period of time. And some countries took that more seriously, I would say, than other countries. And so certain countries had led in only higher-quality tests. But to get through this tough approval, you have to show a lot more clinical studies. So there are actually not many companies that have gotten the full approval. So we believe that although there was some competition, not all of them are going to get full [indiscernible] approval. Now the third one, I'm going to cover on the next slide, actually. That's the GenMark acquisition, which we have closed on April 24, and we're really excited about that. And this is really about the rapid identification of respiratory and blood stream infections. And also the detection of antimicrobial resistance, which is really critical for patients that are severely ill and you don't know what's going on. You want to be able to test loads, whole variety of different pathogens to be able to treat the exact pathogen and not go in with just general antibiotics. Now they have already product on market, mostly sold in the U.S. And given our global reach, we can now move that into other parts of the world. 2 ePlex respiratory pathogen panels with 20 viral antibacterial targets. Here, they also included SARS-CoV-2. And in addition, they have 3 blood culture ID panels. And this is for gram-positive, gram-negative and also fungal pathogens, each with 15 to 30 targets. There's additional panels in the pipeline like gastrointestinal. And what we're doing right now is we're trying to leverage our asset development expertise, our manufacturing expertise and our global market reach to take full advantage of a great product that so far has only -- mostly seen the U.S. market. Now let me talk about 2 very important assays in our portfolio, and that is NT-proBNP and troponin T. We are a global market leader in cardiology, in the diagnostics area. And so far, we've only limited this test for 164 million patients every year. And this is really for people who come in with a chest pain for troponin T who may have a heart attack, so, question, does the person have a heart attack? And for NT-proBNP to really look at heart failure. Now with troponin T, we have done additional clinical studies, and we can now address a much bigger population of patients. One is perioperative risk. So people that are getting surgeries but are not getting cardiac surgeries, but other surgeries, they may have a cardiac events, either during the surgery or after the surgery. And with this test, we can assess prior to the surgery, will this person have a problem. And obviously, it's important information to have. Also, we can now screen the asymptomatic population at risk. With NT-proBNP, we can now target type 2 diabetics who do have a higher risk for cardiovascular disease, and we can also screen for elderly people with risk for atrial fibrillation. So with that, we actually go from focusing on diagnostics and diagnosis to more -- much earlier setting in screening. And this fits very well with our diagnostic strategy to enable better patient care and outcomes while reducing cost of health care by moving much earlier when people are still healthy, you can do something about it rather than when basically the situation has already deteriorated. Now we've launched another assay on our cobas pure, cobas pro platforms, the Elecsys anti-p53. We already have the broadest menu of tests, but we continuously want to push the boundaries, bring new tests onto these platforms because one of the reasons why customers choose our platform is because we have the most tests on that platform. And an important marker here is anti-p53 immunoassay. And this is to aid the diagnosis for esophageal, colorectal and breast cancers. And this is about 2 million deaths every year that are caused through this. And what we're targeting here is the anti-p53 autoantibodies. So our body generates autoantibodies against the mutated form of p53. And with that, again, a screening assay, we can go much earlier, detect, already in healthy patients, if they have developed cancer, and we can do something about it. And this test is then used in combination with other tumor markers or other modalities. Another very exciting advancement in oncology for us is this MMR panel. Now MMR stands for mismatch repair. And the mismatch repair mechanism is one of the repair mechanisms in our body, and we are detecting actually 4 different proteins, the ones in yellow and purple that you see on here. And they're involved in this repair mechanisms. If there is a mutation, a problem, then basically, this repair mechanism doesn't work and results accumulate DNA damage. And as it accumulates the DNA damage, these cells are very good to be targeted with immunotherapies because also the immune cells surrounds, because these cancer cells, because there is such a high mutational burden. And what you're looking here is actually an absence of the signal because then you know one of the proteins doesn't exist, is not functional. And then you know immunotherapies work very well in this setting. Now also in diabetes care, we've made very good progress. And here, we've entered a partnership with Diabeloop, which is a French medtech company that is specialized in algorithms to determine the direct correct dose of insulin delivery. And here -- so now we compare this with this algorithm. And for the solution that you see on here, we've now taken over the responsibility to deliver the first level of contact for customers for this closed-loop solution. And therefore, we're offering all around service for automated insulin delivery for our customers. Now finally, let me say that I'm very much excited of what the team has already done this year with cobas pure, pro and other solutions that we've launched this year, next to all the things that we've done for this pandemic. But I'm even more excited to see what else is going to come. And the team has really worked tirelessly and did a tremendous job to deliver for patients out there. And with that, I hand it over to Alan.

Yes. Thanks, Thomas. 51% sales growth. Okay, not bad. 18 NMEs in late stage, Bill, really amazing. So some comments from my side. Hello to everybody. I hope everybody is safe and healthy. I think still important, I think, Severin, Bill and Thomas have set the stage well. Let me go right to the highlights. Here, we've talked about the sales growth already of 8%, with an 8% increase. Core operating profit with a little bit of a slower momentum but we'll go through that. And then the core EPS growth comes back, if you like, with plus 6% and will clarify that. Cash flow really, really at a good level. And also here, I think, it really helps us certainly from a strategic flexibility point of view. And then really the net financial results, we have been some support here for the core EPS growth, and I will certainly explain the IFRS net income. Good. With that, let's dive into the comprehensive figures. I think really, Thomas and Bill have done a great job in explaining the sales. The composition is still, I think, quite interesting because certainly, I think there was a major boost on the Diagnostics side. And on the other hand, I think we lost a couple of sales on the Pharma side. I think what that came with is, well, as Severin said, we had a different, if you like, ratios for sales. And we had 70% to 71% of the sales in this half came from Pharma. When you look at half year 2020, that was 79%. And I think that explains a little bit what you see later on with the margin, I will get to that. I think when you look really from 8% sales growth to the 4% core operating profit growth, 2 elements will stick out. I will go through them in the P&L. But let me mention them. One is the cost of sales with an increase of CHF1.3 billion, and very clearly driven by Diagnostics, and as Thomas said, driven by the volume growth of 51%, basically no price effect in Diagnostics. So really pure volume growth that also drove the sales growth, if you like. And that was the driver for the cost of sales growth. And then R&D was CHF1.1 billion. And here, certainly Pharma has been a major driver. Core net income went up by 6%. Two things to mention. The net financial result here, the venture fund did a great job and gave us some gains from equity securities, and we kept the tax rate stable. I'll explain that later on, which was, how should I say it, also due to the fact that we had a tax resolution or tax conflict that we really resolved in the first half. You go really further down, you see the IFRS net income get to that. We had really last year a release of a provision here, quite substantial. So that's a base effect and why the momentum comes down in constant rates from 6% to 2% for the IFRS net income. And then you see really the major boost in the operating free cash flow. Thomas has mentioned the GenMark consolidation. Let me clarify these numbers here right away. We closed, as you know, on April 22, it had a sales impact of CHF21 million in the first half. Really, we consolidated from that date on really to half year, so CHF21 million here. And there was a small negative effect on the core operating profit of minus CHF11 million. And all in, the net loss was minus CHF30 million because certainly, I think the amortization kicks in here as well. And when you look really in the finance report, you will find out we had a CHF1.7 billion consideration for the whole deal. And there's now CHF1 billion in goodwill and CHF700 million in the intangible assets. Good with that. Let's go through the core EPS development. Here is the bridge from half year 2020 to half year 2021. And you see really the increase and what has been the contribution from operations. Certainly, we think that this will increase in the second half. You see we had gains on product disposals, roughly CHF354 million in the first half. We are basically at the numbers that we normally have in our numbers. But we'll see what happens in the second half of this year. We will surely be opportunistic once we come to that. Then you see the gains on equity securities with plus CHF164 million and then a couple of other things that brought us to the half year number of CHF10.86. Good. With that, let's go to the P&L. And most of it are explained already. You see now the sales increase, an absolute CHF2.4 billion. And as said, Diagnostics was plus CHF1.3 billion [indiscernible] with plus more than CHF3 billion and Pharma with minus CHF669 million. You see really the royalties & other operating income. This is the gains on product disposals, CHF358 million. You see the cost of sales saving on the Pharma side and an increase -- a significant increase, but well in line with the volumes on the Diagnostics side. M&D, a significant saving of roughly CHF200 million on the Pharma side, here a reasonable increase on the Diagnostics side. You see R&D with a 19% increase, which we expect the dynamic to go down in the second half. So really, it's not like doubling that figure here, and that gives you the full year increase, that won't be the case. Really here a slower momentum in the second half. And then G&A, which is very much driven by the reallocation of costs from other cost lines into G&A, I'm Underlying, I would argue, G&A is pretty flat, leads us to the core operating profit and the increase of CHF425 million. You've seen in the previous slides, Pharma, roughly down by CHF900 million, Diagnostics, up by CHF1.3 billion and group pretty flattish. Good. With that, let's go to the royalties and other operating income, I made a comment here already. You see basically royalty income, out-licensing income, other operating income, pretty, pretty stable. And we have quite some dynamics here. But I think you see really a stabilizing effect. And then you see the income from the disposal of the products that I've mentioned already. And as said, we will see what's going to happen in the second half. Good. With that, let's go to the margins. Severin made a point here already. But let me start on the right-hand side, you see Diagnostics goes up significantly, and that's very much volume driven, and good cost control, as Thomas said. And then you see the Pharma division, I think, very here -- very clearly here, higher investment into R&D, which certainly also brought the margin down. And then overall, I think that's really the mix and the mix of the sales and of the business, if you like, which then comes down to the profit when it comes to Diagnostics and to Pharma. Good. With that, core net financial result, I'm very happy to present that. I think overall, CHF265 million improvement certainly contributed to the momentum of core EPS. Equity securities as at the Roche Venture Fund. Here, I would like to mention we had a pretty strong second half last year. So let's see how that number looks like at year-end, it might be significantly lower. Then you see the net interest income pretty flat, currency, and then interest expenses, once again, we were able to bring the interest expenses down. But basically, when you compare half year to half year with the same debt level of roughly CHF15 billion. Good. With that, let's go to the group tax rate, made a comment here already, 16.9% is a low rate. We had a couple of resolutions of tax disputes last year, which helped us to bring the rate to 16.5%. And what I can say is we also had an impact in the first half of 2021. I expect really the group core tax rate to be at around 18% at year-end. As I said, we had some support here in the first half. Good. With that, to noncore and the IFRS income. And I've mentioned already we had quite a base effect last year. So let me start with the core operating profit. You know the plus 4%, the increase. You see really global restructuring plans, we have a slight increase compared to the last year. Impairment of intangible assets that came down a little bit. And then you see in the legal and environmental line, the positive number in half year 2020, and that has been the release of the provision of a -- of sharing a long-term litigation that we have had, and there was quite significant, and certainly, that didn't show up and reoccur in half year 2021. So really, overall, IFRS operating profit down by 1%. And we had a couple of tax and financial result impacts here, brings to IFRS net income to plus 2% in constant rates and to minus 3% in Swiss francs. Okay. Let's look at cash. And here, really strong development all over the place, you see really in Diagnostics driven by the good business. You see it on the Pharma side. Pharma, I can mention here, we had last year some extension of payment terms for some products in the U.S. Certainly, I think we brought that back and that certainly helped the cash generation, you see really a nice development overall compared to half year 2020. And when you look really at the ingredients of that increase, of CHF3.45 billion in constant rates, you see really where it came from, on one hand, certainly profits. And you see net working capital, and I mentioned that already, really here, we had extended payment terms this year and brought it back to the normal levels this year. The other piece is really here, lower increase in accounts receivables, payables contributed to that and the lower increase in the inventories. When you look at PP&E, an increase of CHF200 million, roughly CHF198 million driven by the manufacturing investments on the Diagnostics side. And then we have lower investments in intangible assets. We had a pretty strong first half last year in doing these kinds of investments. So I think really here, that came down by CHF1.2 billion. But even if you were adjusting that, I think, really overall the cash generation has been on a very, very high level. Good. Group net debt is up versus year-end 2020, and I think that's not surprising because we have paid the dividend of CHF8 billion. On top of that, we went for the GenMark acquisition, which is here under M&A and all, and other transactions, which was really -- GenMark alone was CHF1.7 billion cash out. But what is remarkable though, I would argue is when you look at the net debt level of minus CHF7 billion, compare it with the net debt level last year at the same time and that was CHF8.8 billion. So really, we were able to reduce net debt despite the fact that we paid really the dividend on one hand, increase the dividend. But the other pieces that we had -- also had to cash out for Genmab -- for GenMark. Good. With that, a quick comment on the balance sheet, and this is now a comparison to year-end. And you see cash and marketable securities came down while we paid the dividend. We had the cash out for GenMark. We have the other current assets, here an increase, and that's the inventories, it's the accounts receivables. You see the noncurrent assets. This is really PP&E, roughly CHF1 billion increase, the goodwill, CHF1.4 billion increase and intangible assets, CHF500 million increase. You see the current liability is pretty stable. The noncurrent liability is pretty stable. So it all comes down to the equity, which has increased quite nicely. And now we have an equity ratio of 47%. Good. That leads me to the outlook section, and that starts, as you know, with the currencies. And the currency impact has been lower, as you can see. So evidently, the U.S. dollar strengthened quite a bit. So I think that's really one element that we're seeing. The euro is stabilizing against the Swiss franc and had a positive impact from that. So I think really, overall, you see at half year, a minus 3 percentage points impact on sales, a minus 5 percentage points impact on the core operating profit and a minus 5 percentage point impact on the core EPS. When you look really at what we expect, assuming the June 30, 2021, exchange rates remain stable until year-end 2021, which is certainly pretty unlikely, but it's a model and I think a nice assumption, it gives a feel where we could end. You're seeing that, we think, based on that assumption, that the impact could even be smaller. And you really see at full year with a minus 1 percentage point on sales, core operating profit, minus 3 percentage points. and the core EPS was minus 3 percentage points as well. What is interesting though is if you were taking today's currency rates, you would get to the same impact. Good. With that, let's go to the outlook. And I think Severin has talked about that already. Let me make here a comment, I think we confirm the biosimilar impact of roughly minus CHF4.6 billion that we had before. So I think that's really underlying here. We will surely discuss what's the outlook here for the 2 businesses moving forward. And certainly, what we are assuming is that we grow through the period of biosimilar impact. And with that, I think we are happy to take your questions, and I hand it over to Karl. Yes, thanks.

Thanks a lot. We are in time. I always say there are good problems and bad problems. So we have a good problem because we have so many people online. So over 720. But the good problem also is, a lots of questions. So if we could kindly limit the questions to two per participant. And if we could ask you kindly to keep the answers rather short so that we have a chance to get here through. A - Karl Mahler: So first question is coming from Richard Vosser.

So first question on gantenerumab. Obviously, significant developments from Biogen. Just could you give us an update on your thoughts on how that impacts your program how you're thinking about the accelerated -- a potential accelerated approval pathway? And maybe more importantly, what do you think it will take to get reimbursement in the U.S., we're seeing Biogen get some pushback from payers. Do you think that -- what do you think about the cognition data? How important is that from GRADUATE 1 and 2? And then second question, just on antigen and PCR testing revenues and the development in the second half. Just a flavor of what happened in Q2 to antigen particularly and how you see both of those developing in the second half of '21?

Thanks, Richard. Bill, can you start on gantenerumab?

Sure. Richard, thanks for the questions. So yes, on gantenerumab, I mean we don't really think that the competitive situation has changed the outlook much for gantenerumab. I mean we've had these Phase III studies ongoing for several years now, GRADUATE 1 and 2, when they're completed, will be the largest and longest studies. The primary endpoint is the CDR sum of boxes, which is really the high bar, cognitive endpoint, and we're committed to seeing those studies through. They're going to read out in the second half of next year. Of course, we'll be in dialogue with the regulators about every way to accelerate the process of filing and approvals because we know that people with Alzheimer's are waiting and desperately want better therapies. You asked the question about what will be required for payment. I think -- yes, I think strong cognitive data is required. We're not treating imaging. We're treating patients and patients need a clinical benefit. So we think that's important, and we're certainly committed to seeing these studies through, and yes, delivering the best possible package of data to regulators around the world and also for payers. And I guess I'll hand it over to Thomas.

Yes, sure. Yes. So let me comment first on us, and how I believe that the market will develop. So first, on the PCR side, I mean, we do see that our PCR testing on the automated machines are holding up extremely well. In fact, we've increased our capacity. And with that, we also increased the sales on our automated machines in Q2, where we see a reduction in PCR testing is actually on the more manual instruments, which is the MagNA Pure and the LightCycler. And this is in line with what we said in the past is that people will consolidate on the most automated machines. And it's clear, I mean, in terms of cost per results, 70% of the cost is manual labor. Only 30% of the cost is the test itself, right? So they want to automate because then they will bring down the costs in their system. Now when you look at the markets, you see, for example, in the U.S. compared to beginning of the year, actually, PCR testing is down about 90%. And we don't see the same kind of level of impact simply because we have such a high level of automation, and we have these kind of benefits. Plus as we went into the market, and I've always communicated that we went in with a price that was in line to prior to the pandemic. So we kind of have a buffer, you may say when the situation is that prices are negotiated because there's more PCR volume around. Thus I do believe we have an advantage anyway in total cost of ownership because our instruments are so automated. So I believe it's much more able to hold up in our -- on our side than overall. Now with rapid antigen, it's -- we definitely saw much more of a decline towards the end of the second quarter. And I definitely believe that will be the case, particularly during the summer period when we also have probably less cases. But now with Delta variant, again, you see that some cases are rising. But there's more PCR available. PCR is the gold standard, right? And they will always prefer to do PCR over rapid antigen. And that's why I don't believe that we'll see a similar peak like we had in Q4 last year or Q1 this year when there was just not enough PCR available. Now there is enough PCR available, and I think that will more impact even on the antigen side. So that's kind of the breakdown that I see. And we will see us in the market similar in the antigen. There is no difference because there is hardly any differentiation between the antigen test. Some are better and there are a few that are better. But I would say in terms of -- in the eyes, sometimes of the people using them, they would see that more comparable. So it's more a price competition now around rapid antigen.

I hope we could address your questions, Richard. Simon Baker would be the next.

Two, if I may. Firstly, going back to Pharma R&D. You mentioned that there were some one-off factors there related to the Regeneron antibody. I just wonder if you could give us an idea how much one-off there was in there with a view to the outlook for the second half in 2022? And then secondly, on diagnostics and the potential expansion of troponin T and NT-proBNP testing. That's a very significant market expansion. So I just wonder if you could give us some ideas on what that could mean for pricing, going forward for that test?

Yes. So in terms of the R&D spend on the COVID programs, which included the Ronapreve as well as Actemra and Atea-527. The total spend in the first half was between CHF200 million and CHF300 million in R&D. We think that the number will be significantly less than that in the second half, so maybe less than CHF100 million. So that will help us achieve what we said, which is that the second half R&D spend will be similar to the first half, which means it will be much less growth versus last year because last year, the R&D spend was higher in the second half.

Yes. With regards to your question around proBNP and troponin T, I mean, these markers are already in the market. So we have existing pricing in the market. And so the real opportunity is about the expansion of volume, right? And if we look at tests prior to COVID, if you had a test that was doing, let's say, CHF200 million, that would be for Diagnostics already a blockbuster. So we have really a lot of products that add up to the amount of sales that you see. So I think there's a huge opportunity now and we can work with governments and also health care providers in the different countries to expand that use now based on the clinical data. So it's really exciting to see. But it's never going to be the same level in terms of testing as we've seen with COVID.

Yes. Thank you. Just to complete the picture, we had exactly the same question from Sam Fazeli from Bloomberg on the R&D, questions were just to cluster the questions a bit. Next question would be from -- I hope we could address your questions, Simon. Next question would be from Jo Walton from Crédit Suisse.

Apologies, Karl, can you hear me now?

Yes. Now we can hear you, yes.

I have two questions, please. Firstly, looking at Polivy, we can see that the trajectory of sales at the moment is very modest. Could you tell us what you think the upside will be if POLARIX is positive? And my second question is on marketing spend in pharmaceuticals, fantastic to keep it down 6%, but that's against a presumably a constraint number for last year. You say you could make it down 6% again for the full year. At what point do you think we should expect marketing spend to materially rise given that you have so many new drugs that you need to promote.

Great. Thanks for the questions, Jo. Yes, Polivy, the later line indications that it currently has, there's a limited growth potential. It's a relatively small market in acute lymphoma, and it's sort of shared between Polivy and the CAR-Ts as well as sometimes patients get other chemo containing regimens with Rituxan. So we think really the big potential for Polivy is in the context of first-line DLBCL. And we think the potential worldwide is up to CHF2 billion in sales. So I mean, this is the biggest lymphoma indication, and we're anxiously awaiting the results and hope we have a positive one. In terms of your question about marketing spend, I think we have a lot of launches coming up in the next couple of years. That will put pressure on our ability to continue to reduce the rate of marketing and distribution spend. However, I think the mid- to long-term trend, we believe we can continue to bring it down in the sense that we're increasingly targeting our patients with digital approaches, on-demand approaches, but also we have very differentiated medicines and highly differentiated medicines don't require, certainly on a percentage basis, as much M&D. And so again, we just think that's part of what the world has been demanding is better medicines and lower costs, and we're committed to that as a company, and we think that's a good strategy for the future.

Actually, the same question came from Luisa Hector. Just to complete the picture, that we can cluster the questions. Next one would be Michael Leuchten from UBS. Hope we could address your questions, Jo.

One question for Alan, please. Just going back to the Pharma margin, if I take out the benefit that you had from the disposal gains. So I think that's about 250 basis points on the half year margin, which suggest that you have had an underlying decline because of the R&D spend. If we go back to your commitment to defending the margin, is that -- has that softened? Or is it just phasing as the year progressed -- progresses? And then sorry, just Bill, going back to the gantenerumab question, maybe asking it a little bit more precisely. Your answer, does that mean even if DFT allows a launch without cognition data, Roche would choose not to do that if gantenerumab 1 and 2 -- sorry, if GRADUATE 1 and 2 do not read out on the primary?

Look, well defending the margin, I think that stands. And well, I think -- when you listen, I think there is a clear indication, first of all, on the R&D side, you had to bring the momentum down in the second half. I think we have to work on operations. I mentioned that as well. I think really, Bill said, I think on the cost side, I think we will work as well. I mentioned that for the G&A side as well. So I think that still stands. Don't be dogmatic about that, I think, because there is volatility as we've seen. Also when it comes to the R&D spend, we have seen that, for example, for the Regeneron part but also for Atea. But as I said, it stands.

Yes. And then with respect to the launch scenarios for gantenerumab, I mean, again, I really think what the Alzheimer's community is looking for is a benefit on what matters to patients in their life, and that's not imaging. And I don't think that I've heard from regulators, including the FDA that their content to approve drugs based solely on biomarker data. I don't believe that's the case. And certainly, that's not what we're aiming for. And I'm not going to make any sort of definitive statements about what we would or wouldn't do. I mean, obviously, we have to look at the entirety of the data, but we're very optimistic that we've designed a study that if this MOA and if this medicine is delivering a clinical benefit that we should see it with our study.

And just to preempt the question because it also comes from Tim Anderson and others. Any kind of ongoing discussions with the FDA on a potential filing just because it will come any time later on and I already have it in the written form.

On every program, we have periodic dialogues with regulators, including the FDA, the EMA and others, and gantenerumab is no different. And we will certainly look for opportunities to accelerate the time when we can actually bring this medicine to patients, I think there's a lot to commend gantenerumab, including, again, hopefully, we'll have a positive result in what will be the longest endpoints in the most patients but also a subcutaneous dose, which can be administered at home. That's a big deal. You can imagine millions of patients having to go in every month and get infusions that's got implications for the health care system as well as implications for those people. And so we think it's a really compelling proposition in terms of the profile that we're attempting to deliver. And we'll obviously be working with regulators around the world on innovative approaches to bring it to patients as fast as possible. But beyond that, I really couldn't speculate. And Yes, I don't think that's really in the interest of patients.

Thank you. Sachin -- I hope we could address your questions, Michael, Sachin, you would be next in a line.

Just two. I just want to on gantenerumab and one other, if I may. Sorry, Bill. So based on the prior two questions and answers, it seems clear that you want to launch with cognition data. And just so I understand the acceleration you're discussing is to minimize the time from cognition data to approval. I just want to be super clear that I've understood that correctly and whatever you're discussing with the FDA is around that metric rather than anything else, just to make sure I'm clear on that. The second question is on giredestrant. You flagged a late breaker at ESMO. What do we expect to learn at ESMO that is new and different from ASCO?

Sure. Sachin, I don't know if I'm going to really answer your question. I just think, as said, we will have dialogues with the regulators in the U.S. and Europe and other jurisdictions about ways to quicken the date when a great medicine will be available to patients. And let's hope we have a great medicine, and we'll find out from the studies. What there -- you can imagine there's multiple ways that, that could be done. And we'll be discussing those options in the months ahead with regulators. And if we -- if and when we have something definitive on that, we'll certainly communicate with everyone and keep you posted. Giredestrant will have neoadjuvant data at ESMO. And I think we're also looking forward to the second and third-line study reading out in 2022. But I think there's a couple of things about giredestrant and this whole class of SERDs that are really important. I mean you have -- you really have to look at the side effect profile. I mean these are medicines that people hesitate, even to save patients, because these are women with breast cancer who may not consider themselves patients because they're taking this medicine for years in the case of adjuvant therapy. And so the tolerability profile needs to be really excellent, and that's tolerability profile at the maximum dose because for therapies that act on estrogen and this is a selective estrogen receptor degrader. One of the big issues is not being able to tolerate the full dose. We're very excited with what we've seen. I think many of you know, this is our third SERD molecule. And we really believe we learned a lot from the first 2, and we really like what we see. Also it's going to be important that this medicine be able to be combined with other important therapies in this space. And we've seen really excellent results on that so far in terms of being able to combine our SERD at full dose with other important therapies in early breast cancer. So again, I think we believe this is one of the most important medicines in -- out of 18 that we have in late stage and look forward to sharing more on it at ESMO.

Just to complete the picture, the ESMO data will be Phase II. I mean, ASCO, we showed Phase I. This is the Phase II already. I hope we could address your questions, at least not maybe fully but to the extent we could, Sachin. Richard Parkes would be next.

Firstly, sorry, another one on gantenerumab. The FDA's review of Aduhelm clearly included analysis of multiple clinical programs with various beta amyloid reducing drugs. And they clearly come to a conclusion that there's an association between reduced beta amyloid and clinical improvement. Can you talk about what data you've got in-house correlating beta amyloid reductions with changes in clinical cognitive endpoints with gantenerumab, maybe you've got more analysis in-house than external. And just adding to that, analysts, I think, historically have had this program at a very low probability of success. If you could just let us know what probabilities you've used in your planning assumptions, that would be really helpful. And then second question is, can you discuss what you've assumed in terms of impact of the delta variant on infection rates and testing demand for COVID-19 testing and what's reflected in the guidance? And the reason I ask is that it feels like Diagnostics overall, has recovered faster than you planned for. Pharma now seems -- feels like it's doing a little bit better than you'd planned for. And now it looks like maybe COVID-19 testing demand might be a little bit more long-lived. So why shouldn't I be assuming that your guidance based on your original assumptions is now looking too conservative.

Okay. So on gantenerumab, I think you asked about what data we have correlating the biomarker effect with clinical outcomes. And let's be clear, GRADUATE 1 and 2 are the definitive studies. And they will answer this, and they will be available in 2022, which, given that we're in the back half of 2021, we're talking about very short time frame in pharmaceutical years. So I just -- I know there's lots of questions and curiosity about this. And of course, we're very excited and anticipating -- I mean as to the low probability, I don't think a low probability is appropriate because we wouldn't have gone into the Phase III as if we had a very low probability. We went into it basically based on the totality of the data in the world from the various A-beta lowering medicines and what was seen at low doses and higher doses. And what we know is the plaque clearing -- the strong plaque clearing effect of gantenerumab. We also ran major open-label studies to try to optimize the titration regimen so that we have 100% of patients at full dose within 9 months and to also have a very well-tolerated medicine. And we've got 27-month endpoints on cognition. CDR sum of boxes as a primary endpoint, but multiple other cognitive outcomes that we'll be testing as well as the impact on plaques. And I think it's quite a good lineup, and we look forward to bringing that data to the world next year. And yes, I think as to probabilities, I don't think we really -- we share those. But I don't think we think it's very low. We think it's got a good probability of working and, that's what pioneering medicine is about. Thomas?

Good. Yes, for Diagnostics, let me first say that there's obviously a base effect, right? So if you look at a very strong Q3 and even stronger Q4, then there is a base effect, which will bring down growth in Diagnostics. Then the second piece is, there is -- when you look at sales, there's a volume effect and there's a price effect. And I think specifically, when you look at rapid antigen tests. We do see a very strong price erosion around rapid antigen test. I was mentioning that the differentiation level is not that high. So we'll see that kind of effect because there is a lot of production capacity. I think PCR, we are more defended against this kind of effect simply because we started out at a different position. We have this high level of automation, et cetera. Now you mentioned the delta variant and fourth wave, et cetera, and I do see that, and I do believe that we'll see that. But what you also see is that the curve of the infections and the curve of severe disease and mortality start to differentiate because of the effectiveness of vaccines. So now if people get vaccinated and people are protected, they may not be protected against a source broad, but they will still be protected against being in the hospital and being ventilated. And so the question really is how quickly does the vaccination progress? And more importantly, what other variants are going to come up? And when we look at the pandemic, I mean we are prepared for everything, right? Because we've seen things go and change really almost on a daily basis. So we really had to be agile to address the needs when they arose. And so it's, I think, very hard to predict on how it's going to move. But as long as the vaccine holds up, like it does right now with the Delta variant, I believe, I mean, people will be protected. Severin, do you want to add something?

No, I think you have framed it well. I think bottom line is, there is a significant uncertainty, right? And we don't know exactly how the pandemic will play out. We are not assuming, in our guidance, a severe next wave, which could bring back demand to a level as we have seen it last year. But of course, there are scenarios where indeed, demand for testing will be stronger than we plan at the moment. So we could end up at a higher end of the guidance. But I think, again, what is important to recognize is the base effect, which will be significant as we go into the second half.

Yes, [indiscernible] was asking if you did some modeling actually on the sensitivity of certain vaccines versus the delta variant. And if that could give him any kind of opportunity to be either more optimistic or pessimistic on the outlook of your business because of the sensitivity of certain vaccines to a certain variance, maybe that is a bit too sophisticated. I don't know, but...

No. I mean we have different models. So we modeled a high scenario, we model a low scenario. And so, as Severin mentioned, there is potential upside right, in case things develop in different directions. Now for the world, we don't hope that's going to be the case because if that's going to be the case, it's going to be bad news, right? But -- so we look at those things. So every time there's a new publication, we know exactly the efficacy of Moderna, BioNTech against the delta variants, et cetera, get severe disease mortality. So we track all of that. We know the data from Israel. We know the data from the U.K. So we do all of that, and there are multiple scenarios on how this can play out. And unfortunately, we don't know. Is there going to be another variant that is going to be better in avoiding the vaccine. And we'll see how things will play out, right. But I think PCR, as I mentioned, will hold up more than antigen, and I think that's an important one because we now in the world have more capacity for that.

Thank you. I hope we address most of your questions, Richard. Andrew Baum from Citi would be the next.

A couple of questions, please. Firstly, on the brain shuttle, could you give us the timelines in light of the accelerated approval for adalimumab A-beta reduction. And also to clarify whether it's an IV or an SC drug. I know that you have it in a dose-ranging IV trial? And then second, on Polivy, I understand why your confidence is high, given the enrichment of the trial. Perhaps you could talk to the appetite among community physicians for adopting the drug if it is effective. Some of the feedback we've had is actually there's a very strong desire to adopt it because currently, they transfer high-risk patients for our squared or other regimens to tertiary referral centers, and this would enable them to keep the patients and also keep the revenues. Any thoughts about the commercialization and uptake, particularly in the community testing would be interesting.

Sure. Yes. So on the brain shuttle, it's -- we're in -- I believe we're in dose ranging in humans now, which is encouraging, right? But a lot can go wrong on an early development project. And especially when you're dealing with something that's as sensitive as helping a protein across the blood brain barrier, so I think our PTS is -- that we ascribe is quite low, not because of something particular to this program, but just because -- yes, it's still at a relatively early stage. That being said, I think our hope would be that it would be subcutaneous and that the dose frequency would be rather low because the dose would be low. The whole point I think if I recall right, I think the amount of protein for typical monoclonal antibodies that gets into the brain is something like 0.1%. And so if you have a molecule that's adept at basically crossing the blood brain barrier, and you could take that from 0.1% to 1% or 10%, you're talking like a whole -- well, 2 logs on dose level. So that's very exciting, and we look forward to -- bringing that forward in the years ahead. In terms of Polivy, yes, I think we -- from the early days of Polivy, we were trying to create a medicine that would be well tolerated and would be suitable for use by physicians in the frontline setting, not necessarily in a tertiary care center. That's one of the advantages that we really see with Polivy versus things like CAR-T. And so again, it's really a matter of weeks now until we have a readout and we'll know if we've got a win against DLBCL. And as to adoption, I think if we have a positive study, I think adoption will be rather high because we're talking about a curative setting. And the average age of these patients is not that old, and you've got a shot at a cure for fatal cancer. And so I think even a few percentage points increase in the cure rate would be an important level of improvement. And so let's see what the study brings.

Brain shuttle is actually IV at the moment.

Yes. Well, sorry, the dose-ranging study is IV because that's the easiest way to have a clear PK/PD experiment, but I think our hope would be that it would lead to a sub-Q drug.

Andrew, hope we could address your questions. Peter Welford would be next from Jefferies.

I've got two. The first one is related to gantenerumab, but I won't relate to the point anymore. I am actually relating instead to the diagnostic you're using, I just think you're using a CSF measure to enroll. Could you just talk a little bit about that and whether or not there is a potential, I guess, path to potential approval of that independently of gantenerumab and then sort of how you've used that. And also whether there are any efforts at all to increase the proportion of patients in the gantenerumab studies who may get beta amyloid PET imaging during the study? Or is it just out of that, that maybe you could tell us what fraction of patients will get the PET screening as the study is ongoing during the GRADUATE trials? And then secondly, just quickly on Evrysdi. Curious there, because it looks as though, I guess, quarter-on-quarter, the U.S. growth maybe is a little less than perhaps we anticipated given the strong momentum. But on the hand, incredibly strong start in some other geographies. I wonder if you can talk a little bit about those 2, and particularly ex-U.S., was there any sort of tendering or stocking or anything that we should be aware of? Or is the 2Q numbers, underlying demand that we should be thinking of for the future quarters.

Thomas, do you want to talk about the test?

Sure. And I have to say I'm very impressed how well you know our Diagnostics business. So I'm very pleased about that. So we did get breakthrough device designation for our Alzheimer's test in the U.S. So we are in the process of, hopefully, getting approval also in the U.S. We have approval outside of the U.S. And as you mentioned, these are CSF markers. And outside of the U.S., there's only one other company that actually has that. And that company, I would say, it's more located in one -- one company that's more located in one country. So we are I think only the real global player who has actually an Alzheimer's test. And there is no company yet in the U.S. with approval. So we're working on that. And at the same time, we're working on blood-based markers as well. And it's about PET concordance, obviously, really replacing PETs because, a, PET is expensive, b, the waiting time is very long, and you have to go to very far places to get the PET. Here, you just then take blood, and you can basically do it from anywhere. So we're really excited about that and looking forward to also Pharma coming then with a drug potentially in the future that will really then drive also the testing sales.

Great. Yes. It's been a good partnership. So the question about Evrysdi sales, yes, I think we -- I mean, we did have -- there was a bolus for sure in the U.S. because there was a pent-up demand from patients who -- especially type 2 and type 3 patients who were -- really didn't have a good option. And so we saw that rapid uptake. We do foresee continued demand, continued penetration in the U.S. I think we're at just over 20% market share after 11 months, and we don't believe that's not going to taper off yet. We still believe there's significant more growth, but the pace looks slower because of that bolus effect. And outside the U.S., I mean, we're just seeing a very strong reception because the competitor molecules, they have challenges, right? I mean the need to do the intrathecal injection of the one is certainly a limitation but also the fact that the therapy is confined to the CNS, whereas our therapy is systemic. We think that's an advantage. And then obviously, the gene therapy is only indicated for certain patients, for babies and there are many more patients with Type 2 and Type 3 SMA than there are with type 1 that are living. And so we think that's an important advantage. Let's see, you asked -- sorry, you also asked about PET screening in the study. I thought that we had a sub study. But I think we'd have to check. I'm sure we can get to the percentage of patients that are treated or that are getting PET.

Okay. Next one would be Mark Purcell.

Can you hear me okay?

Yes, we can hear you, yes.

Fantastic. It's Mark Purcell, Morgan Stanley, 2 questions. First one is on Hemlibra. 29% patient share, clearly, very strong performance here. How should we think about peak patient share potential here in terms of do you have any market data to understand sort of what proportion of patients remain satisfied on factor-based therapy. What percentage of patients are maybe still suffering from a fatigue when treated with Hemlibra. And obviously, you've got a next-generation agent here, NXT007. So when should we expect to see Phase I, Phase II data on HEMLIBRA 2.0, given that we should see data on [indiscernible] and fitusiran soon. So that's the first question on Hemlibra. And secondly, on Diagnostics. Thomas, there's clearly an acceleration in routine diagnostic sales growth, impressive increase in Q2. And Bill talked about in on-COVID 95%, 96% of prepandemic levels. So please could you help us understand how far through the catch-up testing phase, do you think we are based on that 31% growth rate in revenues in Q2? And alternatively, the explanation is you're seeing market share gains. So do you believe you're seeing market share gains, if you are, to what extent is that driving growth? And to what extent do you believe those market share gains could be sustainable on a quarter-on-quarter basis going forward.

Great. So in terms of Hemlibra, I think we're currently growing about 1.5 share points per quarter. And we definitely see room for further growth. And again, it's pretty straightforward in terms of the trade-offs. And even patients that are reasonably well controlled on factor, they still have a high burden in terms of treatment burden, having to carry the factor with them. We hear stories from the patients about how they used to -- when they went some where they had 2 suit cases, 1 for their clothes and personal items and the other for their therapies. And with Hemlibra now, many patients are getting 0 bleeds with 1 sub-Q injection a month. And so it's really hard to say how far it goes because we're not hearing, as we do focus groups with patients and things, we're not really hearing the reason why that switching wouldn't continue. It's not like they're saying, oh, there's this disadvantage they have with Hemlibra. So we think we have good room for further growth. With respect to the second gen product, I'd say 2 things for -- in general for second-generation products, there's a very high bar because Hemlibra has established a really strong standard for safety and efficacy. And as you know, for example, in pediatric, which is probably the least confounded by long-term joint disease, we saw up to 90% of patients with 0 bleeds. It's pretty hard to improve on 0 bleeds as a clinical endpoint. And we've been talking earlier about Alzheimer's and the need to treat the patient and not treat the image. And so again, in this case, I think it's treating the patient versus treating lab values. The promise of the second or third generation products would be that it would have a higher level of sort of factor VIII or factor VIII mimicking. And I would say that I believe Chugai's second-generation product would have the highest level. I think their goal would be essentially to achieve normal levels. And so I think if that's important, then we've got a great molecule in the clinic. If it turns out that, that's not important because you just -- you can't really improve on Hemlibra, then we have Hemlibra. So I like our position.

Yes. So to the Diagnostics question, let me first say that Diagnostics is always kind of a leading indicator of what's going to happen later on the Pharma side, because when a patient comes and the first thing that's going to happen is they're going to get diagnosed and only for a certain time delay, then they're going to be also treated. So that's the first part. The second part is, I mean, we do see really phenomenal growth in our base business, routine business. And we always said, let's not take the eye off the ball, right? Well, everything is really busy, let's also make sure that our underlying business is performing well. So we really made sure to have a focus there. And I do believe we are very well positioned versus what else is happening in the markets when it comes to our base business.

Yes. Thanks a lot, Bill, just maybe to give you another opportunity to comment on gantenerumab, sorry for this one. More questions, and I guess you will have a short answer on this one, but I really wanted to complete it. Can you comment on the way on how you want to file? Do you want to go for a breakthrough therapy designation? Do you want to go for an accelerated pathway? Do you want to go for a normal filing pathway? What is the way forward?

Yes. Sure, sure. Yes, I just want to underscore, we have high confidence in an gantenerumab. We've designed these studies with great respect for the previous experience we had with crenezumab and gantenerumab, trying to ensure that we have 100% of patients on the maximum effective dose. We're going to be very patient because the last thing we want to do is interrupt the studies when we know that the time course can be very important. And especially that time where 100% of patients are on the maximum dose after the titration. And we believe we've designed some really excellent studies and that we've got a really great shot at having the best data package in the world, and it could be the best data package for many years to come. And so picture yourself in our position, what would you do with that? And by the way, all that data will be available in 2022. So in terms of speculating and thinking of doing some sort of strange thing that could jeopardize that. It's just not on the table. Now we will have conversations with regulators about innovative approaches to really minimize the time between having data and having an approved medicine. And I think there are many ways that, that could go, and we look forward to those continued discussions. But for now, I think the thing we want to emphasize is we believe we have the best studies, and we intend to run those to their full course and have the best and most clear answer that will enable physicians, patients and payers to have maximum confidence in our medicine for many years to come.

Wimal, I wanted to open your line. it's open please.

Can I first just ask about some of the new interesting trial starts. So TIGIT and Tecentriq combo a very long, the 36-week PDS study and the Kadcyla, Tecentriq combo in breast cancer. So just any color on what drove each of those decisions and particularly any data you can point to that supports these trial starts. And then my second question is on Tecentriq launch in adjuvant lung. Given you filed under [indiscernible], we may not have to wait very long for approval. But given the lack of competition at time of launch, and your view that this is a blockbuster opportunity. How should we be thinking about the ramp in peak sales? Could we expect quite a large contribution in 2022? Any color will be great.

Yes. Great. Great. Let me answer the first question or the last question first. So yes, we've already filed the data. We think the approval for adjuvant use of Tecentriq in lung cancer could come before the end of the year. And as of yet, none of our competitors have had a readout. So we actually are very optimistic about our prospects for helping these patients, make sure that their cancer doesn't recur and look forward to launching it in the coming months. And we think we could have quite a good ramp. So yes, I think 2022, we would definitely see an impact of adjuvant long on Tecentriq. In terms of your questions about why we would start some of these studies, I mean, I think for TIGIT, we have, I believe, it's the only real proof-of-concept study that's been published so far, which is the initial study we did in non-small cell lung cancer with Tecentriq plus and minus TIGIT. And it showed a very strong effect on PFS and on response rate. And it seemed like a very clear signal to us. That's why we do randomized controlled Phase II studies. And that was the basis for initiating the other studies. You'll notice, I think we've been very evidence-based because what we've seen in the earlier studies with TIGIT is that it had -- its efficacy was confined to those patients who benefited from a checkpoint inhibitor. So patients who were PD-L1 positive. And so we've -- I think we've been very prudent in the studies we've chosen, we've gone into settings, either settings where checkpoint inhibitors work for all comers or in the case where the benefit of checkpoint inhibitors is confined to patients that are PD-L1 positive, we're confining the population primarily to PD-L1 positive patients. On Velodrome which is the 36-week treatment interval for the port delivery system with ranibizumab. This was based on the Phase II data where we saw, and we'll have to follow up with the figures, but we saw a high proportion of patients were able to go beyond 6 months. So we ran our pivotal study with the 6-month end point because we thought that was very clear that we would hit that, but we wanted to explore the potential of even longer dosing intervals and Velodrome builds off the proof-of-concept study in that way. Let's see. I think, remind me if I missed any of your questions.

Tecentriq and Kadcyla.

Yes. Well, I just think it's a compelling opportunity where we have in Kadcyla, we have a potent medicine that's been shown in high-risk patients with early breast cancer to have a strong effect. It combines sort of the best of a chemo plus a targeted agent with the HER2 therapy, and we thought it was a reasonable bet that the addition of a checkpoint inhibitor to take the brakes off the immune system could add an additional benefit, similar to what we saw, say, with Avastin in liver cancer with Tecentriq, where we saw a strong signal with the targeted agent and think that combining it with a checkpoint inhibitor would be a reasonable thing to do.

We have three more questions in the line. Keyur, you would be the next one.

Hopefully, you guys can hear me okay. Two nongantenerumab questions, please. One, Bill, 27% R&D as a proportion of sales. You said there's some onetime effect. But just more philosophically, where do you see that number kind of stabilizing over the course of the next 2, 3, 4 years? Just to link to that, Severin, from your perspective, kind of -- how do you see that versus incremental spend from an M&A perspective? And where would you kind of prefer putting more money to work there? Secondly, for Thomas. Thomas, any kind of early data on your usage or throughput of the incremental machines that you have placed during the kind of pandemic. So what I'm trying to get a better sense of is, as kind of COVID-19 testing kind of becomes normalized at some point of time, what should we think of as the underlying growth? And is there a difference in the underlying growth between machines that are sold on kind of a contract basis versus machines that you have placed during the pandemic.

Right. So on the M&A side, perhaps I can start with that question. I don't see any change in our strategy. We focus, as in the past, on bolt-on acquisitions. We focus on technologies or products, which complement our portfolio, and we'll continue to do so. What we have seen and increasingly see is that prices for later-stage assets are very high, and therefore, the hurdles just from an economics point of view are high for us as well. And if we compare such opportunities with our internal opportunities, then often we would have a bias towards progressing our internal opportunities. Now more recently, with this over-proportional spend into R&D, that's, of course, also a consequence of the opportunities we have seen. And as Bill pointed out, we have a record high number of new molecular entities in our late-stage portfolio. We have many indications we are working on, and we had a number of transitions into late-stage development more recently. And that, of course, also both hand-in-hand with an increased R&D spending. Having said that, we see it stabilizing for the second half of this year. So we would expect that R&D ratios will, how shall I say, normalize again? And also, what is important, we want to reallocate money from other areas, right? And that's why you see us working hard on our efficiency in M&D and G&A cost of sales to make those funds available to really progress our portfolio. So overall, no news on the M&A side. And as far as the P&L structure is concerned, we are working hard to keep the margins, but we are reallocating resources from various areas into R&D as our portfolio progresses and as we have interesting opportunities to go forward.

I might just add, Severin, that we're certainly not content with, say, the level of efficiency or productivity in R&D. I mean, we are also taking major efforts across gRED, pRED, Chugai, late-stage to do the same kind of transformation we've been doing in marketing and sales and other areas to make sure that really all our employees are able to make a great contribution every day, clear out bureaucracy, and clear out extra process. And so we think we'll actually be able to extend our R&D dollars further than we would if we weren't doing that. And we've seen that already. In terms of -- for example, if you see the increase, we've tripled -- or sorry, we've doubled the number of NMEs in late stage in the last 3 years, but we haven't taken our expenses up by, I think, we've taken expenses up by 25% or 30%, while we've doubled the NMEs. So we look to additional savings as well. Thomas?

Yes. Thanks for your question. Now to your question regarding the usage of those machines. I can say that there are certain countries in the world where we didn't have a single one, right? So obviously, in those countries, the usage is going to be there, and they're going to continue to use those machines also for other tests, right? And even in countries that did have more PCR instruments, I mean, they really gotten to know first, Roche as really a reliable partner, but also they see the benefit of our systems, the level of automation, the broad menu that we have on our system. So really, systems that don't have those attributes will not persist in the future and systems that have high level of automation and broad menu, they will persist, and this is where all of the testing will be consolidated, even beyond COVID, right, for other infectious diseases, for certain cancer, et cetera. And what we have beyond the high level of automation on our system itself is we have the cobas prime. So even the pre-analytics before it ever gets onto the system. Now for other companies, they would have an army of people just unscrewing caps and pipetting. An error -- a problem because this creates errors, but it's a high cost for that. So this kind of automation is really helping us. So we will get more, simply because they will consolidate. And then the question is, will there be more screening to be done in the future for certain things like certain cancer, where we've been working with governments around the world for the last 10 years. Will this accelerate this opportunity? And I do believe there are opportunities, but some markets are still busy thinking about COVID and others are already talking to us about these kind of opportunities. So let's see how that's going to happen. I mean there are other infectious diseases as well, like hepatitis C and others, where we can really have a great impact. But I believe that if it's going to happen, it's going to happen on our machines.

Two last questions here from -- or 2 participants who have questions. Steve Scala from Cowen.

I have two questions, both for Bill and both follow-up. So on the SERD, Bill, you sounded excited about the data that we'll see at ESMO. Should we assume that excitement implies that the efficacy surpasses the data from other agents with good safety. I'm not sure how else to interpret your excitement. So that's question number one. Question number 2 is apologies on gantenerumab. Your views of the integrity of the GRADUATE program are completely reasonable. But I'm not clear on why GRADUATE can't run their course and you file gantenerumab ASAP anyway, given that the bar is so low. Roche has the regression data now. It seems you're holding Roche to a higher standard than the regulators have put in place. It would seem better to be second to market than third to market. So I assume your fear is that the antibodies are going to be pulled from the market and Roche, rather not be part of that. So can you make any sense of that question and maybe give your perspective.

Sure, Steve. So first, on the SERD, I guess I would just say we -- again, we've been at this a while, and we know the name of the game is trying to get the proper dose levels. And if you look at some of the agents that are available now like fulvestrant, it's a challenge to have significant dosing and to maximize target engagement. And from the abundance of data we have, both animal and human data, we think this molecule looks like the best chance that we can see to really maximize that target engagement in a well-tolerated molecule. It's got good drug properties. Oral, we're just -- it's like all the things you want to see in a therapy that would be taken by millions of women in a routine way. And those things, they will matter. So I'm not sure you'll see everything at ESMO that will make you as confident as we are. But I think you're going to see some new data at ESMO, and you're going to see additional data next year in second and third line. And there'll be additional readouts along the way as we have things like combination data and others that we'll release. In terms of gantenerumab, I wouldn't preclude anything that is an innovative approach to bringing this product to market as fast as possible and preserving the integrity of the studies. So you outlined one scenario. It's just we don't believe -- we believe this topic is supercharged enough already, and it's in the newspaper every day, and we don't believe us adding speculation to that is benefiting to the people who are working tirelessly to help people with Alzheimer's disease. And so I don't think you should view that Roche will take a conservative stance. I think we will take an appropriate stance, and we will do everything possible to advance this medicine as fast as possible. But we're going to do that in collaboration with our investigators and with regulators, and we're not going to play that out in the newspaper.

Okay. Am I still live? Can I ask a follow-up?

Sure, Steve.

Bill, can you confirm it's not already filed?

Sure.

You're confirming it's not already filed?

Yes, yes, yes. It's not already filed.

Good. Last but not least, Marcel Brand from ZKB, then we have to conclude. Marcel? If there is no further question, I would say, let's enjoy the vacation. And thanks for your interest in Roche. Thanks to all of you to be here with us today, and all the best to you. Bye-bye. Have a nice day.

Thanks.

Thank you. Stay safe. Bye-bye.

Thank you. Bye.

Thank you.