Severin Schwan - Chief Executive Officer Dan O'Day - Chief Operating Officer, Pharmaceuticals Roland Diggelmann - Chief Operating Officer, Diagnostics Alan Hippe - Chief Financial Officer and IT Officer

Vincent Meunier - Morgan Stanley Tim Anderson -Sanford Bernstein Alexandra Hauber - UBS Sachin Jain - Bank of America Michael Leuchten - Barclays Luisa Hector - Exane BNP Paribas Keyur Parekh - Goldman Sachs Richard Vosser - JP Morgan Philippe Lanone - Natixis

Ladies and gentlemen, good morning and good afternoon. Welcome to the Roche Third Quarter Results 2014 Conference Call. I'm Stephanie, the Chorus Call operator. I would like to remind you that all participants will be in listen-only mode and the conference is being recorded. After the presentation, there will be a Q&A session. (Operator Instructions) The conference must not be recorded for publication or broadcast. At this time, it's my pleasure to turn over to Dr. Severin Schwan, Chief Executive Officer. Please go ahead, sir.

Good afternoon, and welcome to our Q3 briefing. This is Severin speaking. I'm very pleased with the Q3 results. As you have seen this morning, Group sales are up by 5% at constant currencies, very much driven by our oncology franchise, HER2 up by over 20%, Avastin continuing with a strong growth of 6%, Actemra up in the high double-digits, and also very good growth in professional diagnostics. I should mention that international growth, you know that it's volatile development on the quarterly basis. In fact, it did pick up in the third quarter, now at 8%. It was also a good quarter in terms of the clinical news flow. You've all followed ESMO, really unprecedented data for Perjeta in terms of additional overall survival, good data also on Avastin, cobimetinib and also at earlier stage in cancer immunotherapies. I'm very pleased that we did not only close InterMune transmission, but now we also got Esbriet approved last night by the FDA. So we are now ready to launch during this important medicine to patient who suffer from IPF. If we turn to Page 6, again sales up by 5% on the Group level, 4% in pharma, very strong performance also in diagnostics with 6% growth. On Slide 7, you can see that continuous growth from over the long term. If we turn to Page 8, good growth across the various regions, Europe actually doing fine in a very constrained environment; US stable, up 5% in spite of the Xeloda patent expiry; also Japan solid growth; and as I mentioned beforehand, international markets picking up. Turning to Slide 9, this now reflects the Esbriet again completed this thing to our emerging immunology portfolio. If we look for the remainder of this year in terms of clinical news flow, there is a number of data presentations at the various congresses. What you don't see on Slide 10 is actually San Antonio where we will also present PD-L1 data in triple negative breast cancer. And of course in addition, we expect a read-out of the MARIANNE trial by end of this year. So on that basis, to conclude with Slide 11, we're fully on track to meet the full year guidance with a low to mid single-digit sales growth. We target to be ahead of the sales growth as far as core EPS growth is concerned. And we plan to further increase the dividend. And with this, I hand over to Dan O'Day, Head of the Pharma Division. Dan O'Day: Thank you, Severin. Good afternoon and good morning, everybody. It's a pleasure to present the third quarter results for the Pharma sales. I'll take you right to Slide 14 and you can see the breakout of the pharma sales by region. 4% overall in constant currencies. You can see strong growth in the United States of 5%. This is really led by the HER2 franchise by Xolair, by Avastin. And I'll remind you this is offsetting of course the Xeloda patent expiry. The Valcyte patent expiry or the entrance of generics for Valcyte has still not occurred in the United States. So that has been delayed until next year at this stage, but we will have Valcyte coming in next year. In the EU, we had very strong underlying solid growth on the volume side, overall 3% with some price pressures, demonstrate I think the continued innovative nature of our portfolio doing well in a constrained European environment. We had very good sales of the HER2 franchise there, just getting going there with Kadcyla. I'll come back to that. And we had some good Tamiflu sales in the UK. But all of this is outweighing the price pressure. International, the region is at 3%, up from the half-year. And what we see here is Latin America growing very strongly, some pressures in the Middle East and Russia. And China of course has a tale of two stories really. One is that the long-listed products are being affected by price pressure, but the strategic products excluding Tarceva are growing at mid to higher single-digit. So the underlying business, if you like, is doing well. We do have a local competitor to Tarceva that is applying significant pressure on our sales of Tarceva in China. And we also have in China Tamiflu pandemic order from last year as well. Japan, 7% growth really across the board, very strong growth of the oncology franchise. I would say, if you remember in quarter one and quarter two, we had this effect of the consumption tax and price effects, which kind of distorted the quarter one, quarter two growth. I think in quarter three, you see a good representation of the underlying growth at 8%. I would just make another point that in the quarter three, figures, if you remember, last year we had a 340B effect in quarter three. So without that, the sales growth would be higher. The supply change that we had in the United States that went from general wholesalers to specialty wholesalers had an impact of increased sales in United States in September, and that is expected to be offset or balanced, I should say, in October. So moving forward into quarter four, I guess I just would make sure I draw your attention to the United States figure, because I think we will have this distribution effect being moderated again in the fourth quarter, the full effect of the Xeloda patent expiry in quarter four and then also some increased competition in the field of Lucentis, which I'll get back to as well. Overall, good strong growth across the different regions. Turning to Slide 15, you can see how this plays out from a product standpoint. I'll get back to most of these products. I just want to carve out a couple here, and that Xolair with plus-24% growth, an additional $138 million, really strong growth after the approval of the form of chronic hives by the FDA. Lucentis, you see here, because I won't come back to this, growing at 5%. I can say that Eylea does now have the claim in DME since July. To date, we haven't seen a significant impact on that, but of course we do expect competition in the DME area as we head into the fourth quarter with Lucentis. And the other thing I would say just in terms of general trend is that we're seeing less conversion overall in the market now heading into the fourth quarter and into next year, at least in our opinion, in terms of Avastin conversion to Lucentis in Eylea where we had a bit of that headwind in the past. I would remind you that about 60% of our sales comes from AMD where we have fairly stable market share, 15% from DME and 15% from RVO. And we do have now as of the end of July priority review granted for the new potential indication for Lucentis, which is diabetic retinopathy. It would be the first ever treatment for this disease as well. And then finally, you can see the continued erosion of Pegasys, as expected, with the entrance of the new molecules into the HCV space. We will continue to see this erosion into the fourth quarter and into the early part of next year. And then I think most of the erosion will have occurred in Pegasys in the HCV space. We continue to see opportunity for Pegasys in the emerging market area and in particular in the HBV where we don't have the same competition that we have in HCV. I would just make a point here on MabThera, perhaps coming back to this, but MabThera's growth is a bit confounded by the 340B and also the impact of the change in distribution. But if you remember, last year on the 340B, MabThera was one of the largest products that was affected. And if you look at MabThera just on a quarter three basis, the reported figure was minus-4% with MabThera. If you excluded the 340B impact, it would be plus-7%. And so I just wanted to point that out that in terms of the underlying growth on MabThera, it continues to do well, even though it's highly saturated in many markets. Turning to Slide 16, oncology sales up 5%. I've already talked about the MabThera effect. I would just say that we continue to have in-class competition in Tarceva and the gain in the first-line setting is not completely compensating for the loss in the second-line setting. And I already talked about the specific situation in China. With Zelboraf, I think we're very encouraged, and I'll get back to this, by the recent data obtained with coBRIM, which will provide us with a very competitive combination offering and allow us to again compete in that segment of the melanoma setting. I'm going to dig into the rest of the oncology products in the slides to come, so if we move to Slide 17, starting first with Avastin. Very strong growth now with new approvals in Europe for platinum-resistant ovarian cancer and in the US with cervical cancer brings the total number of cancers that Avastin is approved in to seven across the globe, really strong growth again for the quarter at plus-6%. Avastin grows year-to-date 6% on a constant currency exchange basis. We're really pleased about these two new indications. I think they'll continue to help us drive the Avastin franchise into the future as well. And we see the ongoing growth coming both from these new indications, but also continue penetration in our colorectal cancer, treatment beyond multiple lines, as well as growth in the emerging markets overall. The positive news at ESMO on the HER2 negative breast cancer with TANIA and IMELDA, I'll come back to, but it gives further evidence of Avastin's benefits, particularly treating beyond first-line and the benefit in those follow-on settings. And then finally, I think, in colorectal cancer, again at ESMO, the CALGB data presented there and the subsets continue to support the use of Avastin in the first-line setting for colorectal cancer. Moving to Slide 18 on the HER2 franchise, just another terrific quarter of growth here. We got to 23% growth on the HER2 franchise overall, really unprecedented overall survival benefit, as you've now all seen at ESMO. I hope I didn't overpromise there. I certainly didn't, because with now 16-month additional survival benefit on top of standard of care leads us to a more than five-year survival in the first-line metastatic breast cancer setting. I mean it's just terrific and amazing news for patients and really shows the value of innovation in this setting. We're obviously looking forward with this data to the potential for this regimen also in the adjuvant setting in the years to come. In terms of this overall survival data, we expect the label update at United States in quarter four as well. On Kadcyla in the second line, which is now approved in 63 countries, really starting to get some momentum as well in Europe. Several countries towards the end of this year and early next year are awaiting the pricing and reimbursement. And the product is available to the Cancer Drugs Fund in the UK. So really starting to get some good momentum in Europe there as well. Herceptin, really positive growth overall, 7% year-to-date. The subcutaneous is progressing very well. In fact, in 29 large markets that it is launched, it is enjoying about 20% overall share. And in some markets that have launched either earlier or have healthcare systems that are either more conducive to this, we're seeing shares of about 50% for this at a very early stage of the launch. On a biosimilar side, we don't see any changes. The first biosimilar that has the potential to enter the market here would be in 2016 and that depends on whether or not something is filed in the fourth quarter of this year or the early part of next year. I would just report out on the market shares and then I'll talk a little bit about MARIANNE. In the first-line metastatic breast cancer study for Perjeta in the United States, we're now up to around 55% market share. And in the US, new adjuvant share is up to now very close to 80%, which again shows the value of medicine in the new adjuvant setting. And in Europe, we're already starting to see market shares in the first-line setting of between 30% and 40% in an earlier stage of launch. In Kadcyla in the US, the patient shares in the second-line are up to around 50% and they started the year around 40%, if you remember, and good strong second-line patient share uptakers is being established now in the leading countries in Europe like Germany and the UK where we have shares of between 35% and 40% growth. So I just wanted to make a comment on the MARIANNE trial. I know it's something that you're all interested in, as we are, and we're looking forward to the data that we'll have access to over the end of the year and certainly make a release on the topline data as well and looking to present that at a future conference in 2015 in more detail. But I think MARIANNE is a real opportunity for us across the HER2 franchise. Clearly, the Perjeta data at ESMO shows the benefit of Perjeta plus Herceptin. And if you remember, the MARIANNE trial looks at Kadcyla, Kadcyla plus Perjeta, and then the third arm is the Herceptin plus chemo. So of course it's not a comparison in that trial because of the time we designed the trial. The standard of Perjeta and Herceptin was not established. But having said that, I think there is a variety of interesting outcomes that can come from this. Of course, first of all, we have to wait for the data itself. But I think what it allows is it will give physicians in our opinion a full suite of options and there is a variety of considerations that will come in, first of all, the strength of the data of course, but then secondly, from a patient perspective, the tumor burden, the propensity of the patient to be able to tolerate chemotherapy. And I think there is any number of regimens that can come out of this, including the CLEOPATRA regimen, including Kadcyla plus Perjeta in the first-line. And as you may know, actually Kadcyla is already approved as a single agent in the first-line EU and US for patients who progressed to metastatic breast cancer six months after adjuvant therapy, so for the fast progressors. So exactly what options will play out, we'll have to look at the data itself. But clearly, it allows us to provide a variety of options in the first-line setting depending on the patient need and certainly allows us to continue to keep patients from progressing for a longer period of time depending on the outcome of the data and also the type of patient, whether it's aggressive disease or not. So we'll get back to that, but we think there is a variety of options that will continue to support the HER2 franchise based upon the detailed outcome of the MARIANNE trial results, which we look forward to sharing with you after we have them. So that's on the HER2 franchise. I would come back to now Slide 19 with Actemra, really strong growth, 23.9% growth on a year-to-date basis. You can see the quarter three at 28% with 19% of that coming from the subcu. So the growth in Actemra has and will continue to be driven by the monotherapy patient share uptake and also by the subcutaneous launch. I mean just a point, in the US, the subcu continues to gain a majority of new patients from the subcutaneous patient segment around 80%. So we're really seeing a strong uptake in subcu, particularly in the markets that have had a little bit more time to develop it. Japan, a very strong uptake of the subcu. US also getting a strong uptake. And then Europe, with the approval just recently, we're just beginning to penetrate the subcu setting, but making strong gains in monotherapy there as well. So that's it for the market and products. Let me go to the innovation section, Slide 21. After a very strong ASCO, it was also a very good ESMO for us. I want to talk about some of the cancer results that we had there. But just sticking on Slide 21, I would just make a mention on the Avastin data in renal cancer, which showed, I think, importantly part of our overall immunotherapy strategy, which is we think we have a good differentiated molecule in our PD-L1 molecule for monotherapy. I think one of the real advantages that we have is the ability to combine molecules in the immunotherapy segment. And with now more than 20 molecules in different stages of development in the immunotherapy field alone and then combining those with targeted agents with chemotherapy, we think this is really going to be the game-changer for the future of oncology. And we think we're in a unique position to be able to leverage that. And I think what you saw at ESMO is just the very beginning of those results with the promising early efficacy data in mono and combo with Avastin immuno for cancer and in colorectal cancer showing a good safety profile in combination with Avastin and chemo. So we have much more to do here and obviously many trials ongoing. But it was a good proof-of-concept data point to demonstrate with combination therapy, the promise of even greater efficacy with these molecules, particularly for those patients that aren't responding to monotherapy, which we know is the vast majority of patients in most disease states. We think this is really going to be a differentiator in the future. And then finally on Slide 21, I just wanted to point out again the Avastin data, the CALGB data, which showed the benefits in first-line colorectal cancer again for Avastin. So let me go into a little bit more detail on some of these with Slide 22. I mean this is the data is really terrific news for patients, the delta 15.7 months on top of standard of care. And it really shows the game-changing effect that we're having now in metastatic breast cancer, HER2 positive. On Slide 23, you can see the outcome of the TANIA trial. Again, this is a trial that looked at Avastin use beyond first-line, so in second and third-line. We had statistically significant improvement in second-line PFS, no additional or new safety signals. And this is a trial that we'll expect to have final OS data on in 2015. And then based on that final OS data, we would consider the regulatory steps in the European Union. There is a slide to support the label of Avastin in breast cancer overall. So really good ongoing, continuing data for Avastin. On Slide 24, what you see here is the outcome of the coBRIM trial, which again showed the combination of our MEK and BRAF inhibitor with statistically significant benefits, really close to a 50% reduction in risk of progression or death with the hazard ratio of 0.51. Clearly, the new standard of care, and this was emphasized at ESMO for BRAF-positive melanoma patients is now the MEK and BRAF combination. We also are aware of the fact that the competition also had demonstrated good results at ESMO as well. We think that's overall a very good news for patients. Obviously leading up to ESMO, we weren't exactly sure where the competitive data would come out. But we think we have a very competitive profile with our MEK and our BRAF combination, with a good tolerability and safety profile. Obviously there is a difference between the tolerability and safety profile of our combination versus the competitive combination. But with this very strong PFS data and the establishment of Zelboraf in many markets around the world, we feel as if we're in a very good position to bring this new combination therapy to patients around the world. We've filed this in Europe and in the US. We have received a fast-track designation with filing anticipated before the end of the year. So we look forward to getting this update in our label and getting the MEK inhibitor launched for patients with melanoma. Moving to Slide 25, lots of data. On the immunotherapy program, we wanted to give you an update at this stage on the totality of the immunotherapy program. You can see how extensive it is overall. I would point to the ESMO data in particular. We had good strong follow-up data on Phase II in bladder and our Phase III will be initiated shortly. I already mentioned the promising early combination data. We also had a good safety data for Avastin PD-L1 and FOLFOX in colorectal as well. Overall, we now have six NMEs in the immunotherapy stage and clinical development. And you can see here the many combinations that we've targeted so far across the range of therapies, including some immunodoublets. And you can see the number of new immuno compounds, CSF1R, CEA IL-2, OX-40, CD-40, all making good progress or about to be initiated in Phase I and looking to progress those different mechanisms of action as we know that the cancer immunity cycle is very complex and that at the end of the day, attacking the immunity cycle from a variety of different mechanisms is going to be extremely important. So I would say that we have within our own organization really the most comprehensive immunotherapy portfolio to date and we certainly put a lot of emphasis on that. On 26, you can see the data on bladder. It continues to impress. We're now at 52% for diagnostically-enriched population overall response. As we showed at ESMO, we've got still good sustained durability of response. 19 of 22 responders are continuing to respond, three complete responders in that. The side effect profile is very manageable. And as I said before, we're going to be starting the Phase III shortly and in active discussions with the FDA about our breakthrough therapy designation for this as well and the implications on that. So moving to Slide 27, another study that we're proud to say we've begun in an area of huge unmet medical need, and this is of course dry AMD with lampalizumab. We have now disclosed and announced the trial design here. There'll be two identical randomized studies, CHROMA and SPECTRI. These are trials that have geographic atrophy as the primary endpoint measured at week 48. The trials will continue on. There is a planned evaluation at 96 weeks of two years. And at this two-year time point, we would also be looking at a variety of secondary objectives, which include an evaluation of the patient's visual functions with a variety of measures that we'll be testing on visual function, which is quite important. I'd just remind you again of why we're so excited about this from the Phase II trial and MAHALO, we saw again for patients that had the exploratory biomarker, which will also be exploring in a Phase III, we saw 44% reduction in the progression rate in geographic atrophy for these patients that essentially have no treatment today. So we're looking forward to moving ahead with that trial and the two Phase III trials and seeing the outcome in this very important disease. And then I'll end on Slide 28 with the exciting news around Esbriet. We're really pleased that we got the approval yesterday, just a couple of points on Slide 28 that I want to point out, because here again a disease that has essentially had no therapeutic intervention possibilities, now has an exciting new intervention. What we're trying to show in the left-hand side of this slide is that the median survival overall for the disease is two to three years. I mean it really puts some of the worst of the cancers in some of the lung cancer areas just to give you a frame of reference. So the medical need is huge here. I'm very pleased to say that in our early access program, we have a large number of patients already that have had the opportunity to be treated with Esbriet and we look forward to expanding that reach now of course with the approval. What's important to talk about here, because the disease progresses differently in differently patients is to get the message out to physicians and to the patient community that treating earlier here in the disease that will only progress is extremely important to both quality of life and also the effect on eventual mortality. So we have the sales force prepared and ready to go out. We expect to have the product available for patients in the next two weeks. And one of the key opportunities we have is to educate physicians and patients on targeting the right patients, treat early and being able to really change the course of this disease. So we're very excited. I was with the sales force just a couple of days ago and I know they're looking forward to making a big difference in this disease. I feel we do have a very strong profile of our product with Esbriet. If you remember, in the New England Journal of Medicine, Esbriet demonstrated a pool of mortality benefit at 52 weeks and also had a very favorable side effect profile relevant to the competition. So we're looking forward to bringing this on to the market, and I'm sure you'll probably have some additional questions on them as we get to the Q&A. So let's just move to Slide 30, a quick update on the pipeline since we last spoke on quarter three. Alectinib has had a Phase III start with ALK in non-small-cell lung cancer. And also as I mentioned, we've now initiated the two trials on lampalizumab in geographic atrophy. On Slide 31, a couple of changes, updates since we last spoke. The Avastin glioblastoma has now not been approved in Europe. We continue to have glioblastoma in the US and in many other countries around the world. While we're disappointed with this outcome, the data on glioblastoma in the first-line setting and other lines of setting continues to be strong and we continue to have that as a treatment option for patients in other parts of the world. And at the same time, we had approval for the Avastin platinum-resistant ovarian cancer in EU and also GA101 in the front-line setting in CLL in EU was approved since we last spoke. And then closing on Slide 32, we've got more data to come this year. At the ECNP, we'll be presenting data on our mGlu inhibitor. At the thoracic meeting in Chicago, we'll have updated data on alectinib. This is the Japan Phase I/Phase II study. And then the Avastin lung BEYOND study, which is specifically for the China registration will be presented there as well. And then between those two and Severin mentioned it, we've now publicly disclosed that the additional tumor type that we have shown promising new data in triple negative breast cancer. We will be providing, I would call it, our first encouraging data on triple negative breast cancer at San Antonio and then to be followed up with an ever deeper broader set of data as we go into 2015. So we're looking forward to exposing you to that at San Antonio. And then finally at ASH, we have the Gazyva GREEN study, which is the study that looks at various other chemotherapeutic backbones, particularly important for the US market, where chlorambucil is not as frequently prescribed. And then also additional data on the Japanese trials around ACE910, our hemophilia product, which we continue to remain very excited about. So lots to report on, a very good overall quarter for pharma. And I'd like to turn it over now to Roland to cover Diagnostics.

Thank you, Dan, good afternoon and good morning from my side as well. We've had a good sales progression for the first nine months with a 6% growth in the topline, led by professional diagnostics, which accounts for more than 50% of our sales, doing very well, growing at 8% year-to-date. Molecular diagnostics was 5% growth underlining and excluding the sequencing solutions at 7% and tissue was double-digit growth. So the clinical laboratory unit doing very well with a 7% growth. And then in diabetes care, 1% solid performance with the first nine months here, again operating in a continuously in ongoing volatile market environment. With all these numbers, we can say that we're overall outgrowing the markets. And then moving to the next page, you can see this also in the geographic distribution of the sales. North America was a very strong 6% growth. EMEA and Japan was 3%, each in a more constrained environment. And the emerging markets growing double-digit, Latin America was 12% and 15% in Asia-Pacific. In North America, EMEA, Japan, we see continuous strong uptake for the integrate lab solutions with modular solutions that we can offer. And then in the emerging market setting, that too allows us to create market share, but in addition of course, we benefit from the ongoing investment in healthcare and in infrastructure. And here of course led by China, which continues to do very well with 24% growth for the first nine months. Moving on to the next page with a couple of highlights by segment, I'm not going to go through all the positions here. But on professional diagnostics, the ongoing double-digit growth in immunodiagnostics is certainly noteworthy with 12% year-to-date growth. On diabetes care, we continue to see a good uptake of the more differentiated products, for instance the Accu-Chek Mobile growing at 22%. In molecular diagnostics, you're all aware that we have received the primary screening approval for HPV. Now while this has not been translated into the guidelines, we see still very strong uptake, growth of 57% year-to-date for our HPV assay. And then finally in tissue diagnostics, the advanced staining portfolio, which accounts for the majority of our sales, is doing well with 8% and ongoing good growth in companion diagnostics also with external partners. This is more of a service and development business that's growing 33%, so doing very well here as well. Moving on to the next page and a few highlights from the quarter three in terms of the portfolio, here the AMH, the anti-Müllerian hormone test for fertility, this is the first and fully automated test in this area, a number of advantages for this test around the sensitivity, but also the ease of use and the fact that the test can be used at any day in the menstrual cycle. But I think above this, what's most important is it expands and now leading immunoassay portfolio on the leading menu that we have in our platforms and the test can be performed on any of our immunodiagnostics platform. So very versatile in its use. Moving on to the next page, which is shifting to molecular diagnostics, we're introducing the most advanced PCR system with the cobas 6800 and the cobas 8800 systems with high throughout, the highest automation in the industry, which will complement nicely with cobas 4800. So we're really able to serve all the different laboratory needs from very high to medium or low throughout. We have introduced in the third quarter the system to the blood screening market with the usual multiplex broad panels. I'm not going to go into all the details here of course of all the tests. And we're introducing the same systems for virology in the fourth quarter. So very pleased with these launches and really convinced that they will the change the way molecular diagnostics are performed, especially in the largest labs, which continue to drive efficiency. Moving on to the next page, which is dedicated to diabetes care. And as you've seen despite the challenging environment and despite the price pressure, we see continued potential for differentiated and innovative solutions, which we introduced two here this year so far. On the left side in blood glucose monitoring, we have the Connect system, which is a wireless connection between the blood glucose meter and iPhone and a PC. We call it the Connect system and it is based on the already introduced, well implemented Aviva and Performa systems. What it does? We believe it has the potential to improve therapy outcomes, easy data management, better accuracy in insulin calculations. It uses the existing bolus advisor that we have introduced on the systems. And of course, it also will enhance the communication between healthcare professionals and the actual users. On the right side, you can see on the diabetic therapy systems a new system called Accu-Chek Insight, which is a combined meter and pump system with a two-way Bluetooth communication, remote control with again a proven bolus advisor and a very discrete handling, which is very important for patients on insulin, so a great system here. And with that, I'm turning on to the next page. I would like to highlight two recent examples of the actuality in those of the emergencies of the value of diagnostics and what it can contribute commercially, but then also to society beyond that. So the first one is the Global Access Program, which we have announced on September 25, which is an exclusive partnership with United Nations and the United Nations AIDS program, and you can see there also is prominent other institutions such as the Clinton Foundation, the PEPFAR, the US President Emergency Plan for AIDS relief, also The Global Fund to eradicate AIDS, TB and malaria. There is a special access and pricing program with emphasis on HIV viral load testing. You can see the goals on the right side, which are aiming at diagnosing 90% of the population living with HIV and the people should receive sustained anti-retroviral therapy. And of these again, 90% should have a durable viral suppression, and this is targeted by the United Nations for 2020. Very excited to be able to contribute here. I think what's also interesting and exciting here is the test is running on our installed systems, which is the proven AmpliPrep and cobas TaqMan systems. So we're using an installed capacity and we're also reaching out to a healthcare population that has already been trained in many instances on how to use those tests. We have an education training center in South Africa that supports us in doing this. So this is the first one and then turning to the next page, just an opportunity here to just share with you that we do have an Ebola test. We distribute on our PCR systems on the Lightcycler and the cobas z 480. This is currently used or approved, I should say, for research applications only. However, we are in negotiations for submission for emergency use authorization by the FDA and also with WHO for pre-qualification of the test. What's important here is of course I would hope that we can contribute to resolving or contributing to helping in the global healthcare crisis that we're experiencing to date. And that takes me to the last page, which is on the key launches. I won't take you through all the products and the launches that we have, but you continue to see our commitment to providing high-testing efficiency on the instruments and the novel assays and markers on the testing side. So again, good portfolio delivery and most of the portfolio right on the launch plans. And with that, I'd like to turn over to Alan on the financials. Thank you.

Thanks, Roland. This is Alan speaking very brief comments on the finance side. When you go to Slide 44, what I'm going to talk about is about the currency impact as well as the acquisition of InterMune, how we did the financing. On 45, you see the information we have. What I would like to emphasize is the currency impact. And you see it here, this light rose bar, so to say, with a minus-$1.711 billion, which brought the growth rate down from the cost and currencies of plus-5% to pretty much 0%. Nevertheless, I think that is less than we've had at half-year with minus-6%. And when you look really at the impact, where do they come from, the major impact is from the US dollar, which accounts roughly for a third of the whole impact and then the Japanese yen was about 17% and then Argentina and Brazil combined, so Latin America was roughly 15% of the impact. So you see three major currencies are pretty much explaining that impact. With that, let's go to 46. On 46, a quick update on the acquisition of InterMune, a couple of details here, as you see, the tender offer and the second step merger. So really 79% of all shares were tendered when we did the second step merger on September 29th. And we did pretty much in parallel the bond issuance on September 22nd. And we have raised $5.7 billion. Impact on the outlook, nothing has changed here. The financial impact is expected to be neutral to core EPS in 2015 and accretive from 2016 onwards and we don't expect any material impact from the transaction in 2014 to sales and core earnings outlook. Nevertheless, as you know, it's excluded from the guidance. so let me go to the next page, which is 47, and let me talk quickly about the InterMune financing. As I have said, we have raised $5.75 billion, which was beyond what we had targeted on. We saw about $5 billion might be the right number, but the order book was oversubscribed and we took advantage here. So the market was very receptive. And you also can see that from the pricing, the spreads are exactly what we wanted to be and we had a two-step approach here. So really bringing in the pricing that we had in mind. Well, treasuries jumped a little bit before we did the transaction. Nevertheless, we have achieved everything we wanted to achieve and it's really a benchmark deal for a AA rating. When you look at the Roche yield and certainly the weighted average, you see here it's 2.286%, I think pretty favorable enough for the maturities that we have in here. So that provides us really a very attractive financing going forward. And with that, let's go to the next page and look at the debt maturity profile on 48. And first let me emphasize that 72% of the Genentech related debt has been repayed. So we have from that transaction $13 billion left. And you see that with the hatched bars we have the InterMune financing now coming on top. From a maturity profile, it's not perfect. I told you at mid-February, we would have loved to see more in 2018 and 2022, but the market has standardized, as you know, in 144A. But I think it clearly outlines that we have things to work on like 2019. So there are still opportunities to improve our structure here. With that, let's go to 49, and 49 is one of the standard slides that you already know, and it's dealing with the currency impact and assuming that the rates at 30th of September remain stable until the end of 2014, the impact we should see at full year on sales should be minus-3 percentage points, on core operating profits minus-5 percentage points and on core EPS minus-5 percentage points. So I think we see a certain mitigating of the currency impact based on the conventions that we have applied here. We will see what happens, especially to the US dollar. But I think with certain developments that we're seeing today, looks like the other things are remaining perhaps a little bit stable at the end of the year from a currency point of view. And I think my last slide on 50 is the outlook. Severin has talked about that already and supported the outlook. So we are confident that we can realize the guidance at year-end. And with that, thanks a lot and we're happy to take your questions.

(Operator Instructions) First question from Mr. Vincent Meunier, Morgan Stanley. Vincent Meunier - Morgan Stanley: The first one is on Esbriet. Can you please comment on the recognition of the mortality benefits in the label you've obtained? And also, to what extent the function would be relevant in order to differentiate Esbriet versus (inaudible)? The second question on MARIANNE, following the publication of this label data from CLEOPATRA, is it fair to assume that now Perjeta and Herceptin is becoming the new standard at least till the overall survival data of the MARIANNE data is published? Or in other words, is the PFS data from MARIANNE relevant and maybe we should just wait for the overall survival data of MARIANNE? And maybe the last question on PD-L1. Can you give us an update on the discussions with the FDA on the possibility to file in bladder cancer on the back of Phase I or Phase II data earlier than the 2016 filing? Dan O'Day: So let me get to your questions, first on Esbriet. Again, obviously we just received the approval yesterday and we're looking at all of the data in question that you asked. But let me just give you a little bit of insight into the mortality data and some of the differences between the label and some of the other data that's been discussed out there in the community, for instance, at the respiratory meeting in May. Now the New England Journal of Medicine article looked at the mortality benefit for Esbriet at 52 weeks. And in the package insert, the FDA included the couple of markers from the end of study, which in our case was 120 weeks. So I think that's one of the key differences there between the two data points. In the New England Journal of Medicine article at 52 weeks, in the pooled analysis of the two capacities, we saw significant mortality benefit. And I think that has been what has been discussed within the thought leader community and one of the very exciting aspects that Esbriet brings to these patients. So I hope that helps put some of the context into the perhaps two different mortality pieces that may be looking at. And then secondly in terms of liver function, clearly that's again something we'll be looking at on a comparative basis, but I do believe that the liver function profile obviously very important for patients and yet to be seen on the difference on the competitive side. But I think it's well represented in the package insert in terms of the liver function capability and also how to manage that. On the MARIANNE trial, I'd try to frame it a bit. I mean it's extremely exciting news on Perjeta and Herceptin. I think that's only a benefit to patients and to the overall HER2 franchise at Roche. Clearly we have to see the results of MARIANNE. But to your point about PFS versus OS, I think PFS will be very meaningful in this context in terms of looking at how to combine these. And again, I believe there will be different options for different patients depending on the outcome of the results. I think there will be some patients for which the Perjeta and Herceptin regimen is the right regimen, particularly with the chemotherapeutic background. It could be patients with high tumor burden. We want a more immediate response. And there's going to be other patients and again it depends on the results where we're hopeful that the Kadcyla plus Perjeta data will demonstrate a significant benefit and also in chemo-free regimen. And then we also have to consider how physicians will look at first-line versus second-line versus third-line and they'll line up their options. Again, I don't want to get too far ahead of the data, but I think the point is that we see with this data that patients are living longer and that as they progress, there will be multiple different options that patients will have and physicians will have to be able to choose from. But to your specific question, I do see PFS as being a very relevant indicator for how physicians and patients will make choices on this. And then finally PD-L1, the discussions are ongoing right now with the FDA. So I can't give you any more color that that at this stage. Very good discussions with the FDA around the bladder cancer and the entire program, so we'll update you more as we get more on that. But clearly, we would be pursuing the fastest, safest way to get this product to market and bladder presents a very interesting opportunity for an early market entrant. We continue to make very good progress on our lung cancer trials. And as you know, we'll be getting the data in-house by the end of this year and the beginning of next year, but it'll be instructive to assist us with the label enabling trials also around lung cancer.

Next question from Mr. Tim Anderson, Sanford Bernstein. Tim Anderson -Sanford Bernstein: On the PD-L1, in FIR trial, can you just update us on the timing of having those results and also just describe the possibility of using that as the basis of registration in first half '15 in lung? And then second question is can you talk about kind of a commercial case for a MEK and BRAF relative to a PD1 or PD-L1? Let's assume that with the pharma, you still don't get durable benefit like you do with the PD1 and I know the PD1 right now that's on the market in the US is only approved in second-line. But ignoring that for the moment, it just makes me wonder what a preferred regimen is going to be here. And then last question is on crenezumab versus gantenerumab, is it realistic that you'll really take both of them through full Phase III development or would it more likely be that you kind of choose one product over the other and can talk about what the gating factor is from here on whether move crenezumab in the Phase III? Dan O'Day: First of all, on PD-L1, FIR timing, remember this is a single-arm trial, it's not a randomized trial. And the whole objective around FIR was to allow us to have data to be able to inform us better on the other trials associated with the non-small-cell lung cancer. So both FIR, which is a single-arm; and POPLAR, which is a double-arm trial, were designed to assist us both in the statistical analysis plans for the labeling trial of OAK, but also to take a keener look at the diagnostic itself and also potentially the cut-off associated with the diagnostics. So that's the primary objective. It's not our base case. For POPLAR would be approval trials. Obviously we'll look at the data like we would with any trial, share with the agency and have discussions around that. But I want to caution as to little bit on the expectations there, so that you know what our base case is. In terms of the MEK/BRAF combination, I think you bring up a very good point. The first thing to be said is it's terrific news for patients with melanoma that haven't had anything for years. And I think that the two will be used clearly with patients in different ways. So as you rightly pointed out, with the PD1, that does come in at a bit more on the second-line. But having said that, I think from a physician standpoint, many physicians I've spoken to, I think they could use these in different sequences and different orders. Because many melanoma patients are diagnosed quite late or in the later stages of the disease and we know that the MEK/BRAF combination has a fairly rapid effect in these patients, we can see a scenario for those types of patients where our MEK/BRAF combination will be the first choice to stabilize the disease, understanding that even though we have longer survival, at some point in time they progress, and then adding the potential immunotherapy compound to get a more sustainable, durable response in those patients that would respond to a PD1 immunotherapy. There could be other patients for which the disease is at a much earlier stage and there's time to see whether or not a patient responds to immunotherapy first and then potentially using the BRAF and MEK combination in a secondary. I am speaking about all of this for BRAF-positive melanoma patients. Obviously for BRAF-negative patients, that's a different scenario. But I do believe that there will be obviously a use of both of these in different sequences, in different orders depending on the patient type. And we have to remember that really rapid control of disease that we see, I mean you see kind of overnight changes with these patients when they're BRAF-positive and they go on to Zelboraf or Zelboraf and MEK combination. And that has to be taken into account. And then finally on the whole anti-amyloid programs at Roche, obviously gantenerumab is our lead program. We're into the enrollment on our mild and moderate study, the Prodromal study continues. With crenezumab, as I communicated at the half-year, we're still in the stage of evaluating the encouraging results, I would say, from ABBY and BLAZE. As I indicated, we have a number of things to look at, including discussions with regulatory authorities, which we're in the process of doing, that will enable us to be able to make the best decision on crenezumab, which we hope to make before the end of the year at this stage. And we think there's a variety of different options, a variety of different ways forward. There's best-in-class type options for this. And all of that data is now being analyzed as we move forward. What's for sure is that given the unmet medical need in this area, there's more than ample space for multiple programs within the field of Alzheimer's. And the question comes down to the data, how we evaluate that data. And we'll make that decision a little bit later this year at this stage.

Next question from Alexandra Hauber, UBS. Alexandra Hauber - UBS: Firstly on MARIANNE, the read-out has been pushed out a couple of times, because it's event-driven. Do you actually know from the safety data monitoring that we are close or is there a chance that it is going to be pushed out even beyond late this year? Secondly, on Avastin in Europe, we've actually seen a very consistent decelerating trend with about 17% local currency growth in the third quarter last year, then 9%, then 8%, then 2% and now 1%. What is causing that deceleration and where do we go from here? The third question is on ACE910, in the appendix, you mentioned an extension study, which is started. Could you maybe give us some color on that, how many patients we're talking, whether that includes Caucasian patients and also perhaps in which dose? And the final question is actually on tax. With InterMune, assuming it is successful as the said internal InterMune documents suggest, your US pharma sales clip surely towards 50% of pharma sales. And what does that mean for the corporate tax rate? Where does that go in let's say 2018 and is there anything to mitigate that? And just also on tax, I just noticed that the two advisors to InterMune certainly took a very different view on what the tax rates for InterMune would be, and I just would like to know which one was right in your view. Dan O'Day: So first thing on the MARIANNE trial, I think that's pretty straightforward. It is now closed. I think we will be reading that out by the end of the year. In terms of Avastin, first of all, Avastin overall growth rate 6% through the third quarter, very strong, new indications coming out. In your statement on Europe, I mean when you look at the data itself, it's overall on an absolute basis quite strong quarter-to-quarter is what I would articulate. There are some price effects in Europe that are taking place at different times that could be affecting slightly the impact quarter-to-quarter, but the underlying demand for Avastin in Europe is extremely strong. And remember, we just had new indications now launching new data, for instance with CALGB, where particularly in Europe, we had a bit more pressure than other places that reaffirms Avastin's role in the front-line setting now with the ovarian, with the expected cervical cancer. So I want to be very clear on Avastin that we don't see a trend of a decline with Avastin at this stage. In fact, it's completely the opposite. And then finally with ACE910, the extension study of course will be very similar to the population in the initial study, which was predominantly Japanese. And so the extension study will also be predominantly Japanese. There are some Caucasian patients in there. And I think when we present the data at ASH, you'll be able to see that in a bit more detail.

I think we're on our way to really integrate InterMune and to consolidating the company. I think it's very clear that evidently IT for Europe is in Europe, IT for the US is in the US, and yet you make the right point, Alexandra, because in 2018, the sales will be quite significant for our Group. And therefore, the profits will be quite significant that we get from that, also molecules, just to put on the table yet. I think when I look at it from today's point of view and 2018 is a bit out there. Yes, I think it could be that the tax rate has a little bit of an uptick, but I don't expect a significant change to happen. Certainly we will optimize it as we go and we'll look at different things here, but seems to be manageable.

Next question from Sachin Jain, Bank of America. Sachin Jain - Bank of America: Just firstly on Esbriet, could you confirm the pool's mortality analysis you have in NEGM is available to be used in the marketing materials, given it's not on the label? And could you clarify why the FDA had an objection to that NEGM analysis? Second question in on triple negative. Do you believe you've got sufficient data at this stage to progress this to a registration study? And if so, when could that be? And are there any specific combinations you target in this tumor type, given you've cited in your intro significant interesting combinations? And the final question is on Gazyva and the ongoing NHL studies, [ph] GO and GAMBLIN. Could you just tell us how many interims you've been through on each of those studies and when the last ones were? My understanding was you might have some recent interim analyses on both of those studies. Dan O'Day: On the Esbriet side, obviously we're just getting the data in on that. The pool data can clearly be looked at, at the 52-week mark and at the end of study. So I think both are clearly accurate. In terms of exactly the promotional stance on that, again I think also due to some of the competitive nature of the market out there, I would leave it at that statement that it's 52 and end of study. They're both ways to look at that. And we're just digesting that and deciding how to move ahead of course into the marketplace. Secondly in terms of the triple negative breast cancer, this is Phase I data. Remember that we'll be looking at here that it's a bit too premature to determine exactly how we'll move ahead or if we'll move ahead to a label enabling trial. My expectation is we want to progress this into a Phase II. But let's first look at the data that we'll expose you at ASH and then even more data into next year. And then as the program develops, I'd be able to update you a bit more on the plans for that as we move forward. And then finally on Gazyva, the interims so far have been positive. I'd just update you a little bit on the expected dates for the trial. Of course the aggressive non-Hodgkin's lymphoma or DLBCL, that one we would expect to read out in '15 at this stage. And then the two interim trials, there's a relapse refractory and then the front-line. The relapse refractory would be the next one to read out, which should be in '15, '16 timeframe, so a little bit of flexibility on that. And then I would say on the front-line, we're talking about '17 still. But the read-outs to date and all the interims have been positive. Sachin Jain - Bank of America: You mentioned ASH within your PD-L1. I just wanted to check whether there was an additional disclosure or slip of the tongue. Dan O'Day: Sorry, San Antonio. It'll be San Antonio, I apologize. That was a slip of the tongue.

Next question from Michael Leuchten, Barclays. Michael Leuchten - Barclays: Firstly, would you be so kind to break down the 5% CR growth in the US in pharmaceuticals into volume and price growth, if you can? Second question, pharmaceuticals, you have changed some of your distribution in the US. I just wondered if you could talk about the rationale both from a safety perspective, but also from the rebate perspective, whether that helps you from rebating side. And then lastly on lampalizumab, where are you with the discussions on GA as an end point with the FDA? I appreciate you made comments of one of your endpoints, so 48-week endpoint in the Phase III trials. But you're also looking at two-year endpoints in the secondary endpoint. Just wondered what that means for the regulators, whether they're going to be okay with GA only or whether you need vision improved as well? Dan O'Day: On the volume price in the US, I don't have all the detailed figures right in front of me. But suffice it to say, that's largely driven by volume. If you remember, particularly on our cancer products and the cancer portfolio, price effect is really minimal these days because of the nature of the US system. On some of the other, if you like, older, longer-listed products, there is an ability to increase prices, but I would say really what you see in this 5% is largely, largely volume-driven. In terms of your question around the distribution change in the United States, as you know, with the products we've recently launched, Perjeta, Kadcyla, also Gazyva, they have been launched through the specialty distribution channels. I mean it takes the number of distribution channels down from 40 or 50 down to around six. So it really allows us to much better control our inventory, make sure we control also the supply to patients, the cold chain, the reliability of our products from a variety of different aspects. And we think it's clearly in the best interest at the end of the day of making sure we get the right product to the right patient at the right time. In terms of the benefit, we do get some efficiency benefits from that. We get some efficiency benefits obviously from working through fewer wholesalers both in terms of the supply chain management and also in terms of potentially some discounting as well. On lampalizumab, I would say on the geographic atrophy, absolutely, I mean the discussions and the design of the Phase III trial was closely discussed with the FDA and the geographic atrophy endpoint is something that has been agreed at the one-year endpoint. We think that the two-year endpoint and the second measures will also be important, I think, for patients and clinicians to know and could also be important outside the United States as well.

Next question from Luisa Hector, Exane BNP Paribas. Luisa Hector - Exane BNP Paribas: The first question, just looking at the HER2 franchise and the strength in the sales, thinking a little bit about profitability, can you comment whether you're adding any rep ex-US now that you have the various options? And could you talk about how the subcut formulation fits in, whether it fits nicely with using Perjeta, whether there's any conflict there? And then the second question on biomarkers, if we look at the lebrikizumab and the lampalizumab studies where you have biomarkers, as you go into the Phase III, both studies are taking all comers. And I just wanted to clarify why. Is this a regulatory requirement? Is it because the biomarkers are not fully accepted yet, because initially we're thinking biomarkers could allow for kind of smaller and faster studies, but certainly outside of country, that doesn't seem to be the case. So can you give any color on that please? Dan O'Day: In terms of the rep, so the sales force throughout the world, there is no universal answer to that, I would say. But I would say as we launch new products into the metastatic setting and then eventually into the adjuvant setting, we get to points in countries where we do need to add new sales force as we want to get the messaging right. We want to get the messaging out there to our customer base. And frankly, it just becomes a rather large portfolio to carry when you consider all the data in second-line, let's say, for Kadcyla, third-line and beyond and then also trying to emphasize the first-line messages on Perjeta and also keep the emphasis on Herceptin in the adjuvant setting. So the answer is yes, we do incrementally add sales forces around the world to make sure that we can get the message across, but we also of course in these channels have efficiencies. And whenever we can, we try to look at efficiencies. So that's about as much color competitively that I can give you on that. But we are looking to expand our sales force as we have these new indications. On the subcut, I think you raised a really good question. Remember the vast majority of Herceptin's use is in the adjuvant setting today. And when we think about the biosimilars coming in 2016 potentially timeframe and then also the timing of our follow-up adjuvant studies with both Perjeta and Kadcyla, I think it's very important to have an opportunity to continue to make sure that we protect our franchise and make sure patients receive Herceptin. The subcu formulation has a big play in the adjuvant setting in terms of the biosimilar. It allows us to bridge between them. In the metastatic setting where anyway you are getting Perjeta or getting Kadcyla, I think that the benefit equation is less. And so in some countries, we're seeing total conversions on both the metastatic and adjuvant setting, in fact many countries. But I think it is a strategy that allows us, I think, to put our best foot forward in both the metastatic and the adjuvant setting over time. So I guess I just have that additional consideration for you. You raised some very good points on biosimilar with lebrikizumab, lampalizumab. First of all, why are we doing all comer studies? Again, these are large patient populations where we had very compelling data in both cases, but just take lampalizumab, for instance. In the all comers, we had a 20% effect. And in the diagnostically-enriched, we had a 44% effect. So the first thing is that we're seeing responses, of course, in non-diagnostically-enriched population. And therefore from a regulatory perspective, but also from a patient benefit and ethical perspective, we feel it's important to have all-comer trials. And that's maybe the same case in lebrikizumab, although lebrikizumab, I think, the response rate in the non-periostin-positive on a proportional basis is probably less. So I think what we'll see in areas outside of cancer where the fundamental understanding, the biologic nature of the disease is by definition less well-understood. We will see trials that are all-comers. So we may not get the complete efficiency in terms of how we run our trial programs. However, I think when we get to the market and we have a look at both determining which patients are most likely to respond to help physicians understand how best to prescribe these products and then very importantly from a payer perspective when we start to consider reimbursement of these programs in larger populations, we think the fact that we could demonstrate significant additional clinical benefit in a diagnostically-enriched populations could help many markets around the world to create access for these medicines where in an all commerce population, they may be more resistant. So I think it's a part of our total strategy. Of course wherever we can, we try to have efficiencies on the clinical trial design and speed, but at the same time, that isn't going to always be the case. Luisa Hector - Exane BNP Paribas: And, Don, are the regulators pushing for the all-comers as well? Dan O'Day: Yeah, of course. Just take that lampalizumab example, yeah, the regulators would not have wanted us to diagnostically-enriched when you have that type of response rate in an all-comers population. That's very clear.

Next question from Keyur Parekh from Goldman Sachs. Keyur Parekh - Goldman Sachs: I have three pipeline questions, if I may, please. First, if you can just give us some sense for how you're thinking about Esbriet beyond IPF. You now kind of closed this transaction for about three to four weeks, if you could just give us a sense of that, that would be great. Secondly, as it relates to the PD-L1 either mono or combination in hematology or broader blood cancers, can you just help us think about when we might start to see some data from Roche perspective? Obviously we might see some data for the PD1 at ASH. And then lastly on the severe depression program or the adjuvant depression programs, you got two programs that you have data for in-house, one is decoglurant and then the second one for the basimglurant. Can you just help us think about how we should think about the difference between the two and how we should think about Roche's appetite for taking either one or two of those into a large-scale Phase III or would you look to partner those compounds out? Dan O'Day: First of all on Esbriet, I think it's really important to say that the InterMune acquisition was really based on Esbriet in the field of IPF. The most important and key goal that we have right now is to make Esbriet a success in the United States to expand this success outside the United States. And we're 100% focused on that. Having said that and I would say opportunistically, we are of course looking at Esbriet and potentially other compounds and molecules within InterMune and within our own portfolio in the general field of fibrotic diseases. So it's a bit premature to get too detailed on this. I mean InterMune had some programs looking at Esbriet in scleroderma. We'll continue to look at those. We've got programs in our portfolio where we're looking both in the early research stage at fibrotic diseases, but then also products like lebrikizumab, for instance, where we have a Phase II program going on right now in IPF. And it's a potential mechanism as well for (inaudible) either alone or in combination with Esbriet. As you know, IPF and actually most fibrotic diseases are pretty heterogeneous diseases. We'll also be looking at any possibility to look at biomarker hypothesis in these diseases. But usually with these complex heterogeneous diseases, there's different mechanisms of action. There's possibility for combination therapy. So all those things, I think, are things we'll explore as we continue with the program. And as we get deeper into the science here and deeper into the integration into our own programs, I'm more than happy to update you on that. On the PD-L1, yeah, additional data. We've got the triple negative breast cancer coming up at San Antonio. We've got the FIR and POPLAR studies end of this year, the beginning of next year. I think beyond that, obviously all of our programs including the ongoing bladder programs and renal cell with Avastin, I don't want to be very specific about timelines next year, but you can certainly expect a complete update at ASCO. And when we get into the first quarter of next year, I can potentially be a little bit more specific on some of those programs. But it is really a massive program now at Roche and Genentech and something I spend a lot of my time on, making sure that we're progressing. And as the data becomes available and as is meaningful, we'll certainly be sharing it with you. And then finally on the depression products, similar to the crenezumab and gantenerumab, big unmet medical needs, high-risk, high-reward potential programs. And specifically, the most advanced program here is the mGlu, which we're going to get some insight into at ENDO. We're assessing the data on that program in terms of how to move ahead with this. We would never exclude options to look at partnering any compound in our portfolio if we think that's in the best interest either from an expertise standpoint or also sharing of the risk standpoint. I don't want to say more on that at this stage other than I think those are all things that we're looking at and exploring. And after you see some of the data on mGlu, I think we could be in a position to discuss and dialogue more about our progress and our next steps in the depression program.

We'll now take the question from Mr. Richard Vosser, JP Morgan. Richard Vosser - JP Morgan: Just firstly a follow-up on lampalizumab, you highlighted the FDA is looking at the geographic atrophy endpoint. So with AMD, they also wanted two-year safety. So just to clarify whether you can file after the one-year data or you need two years for the filing on lampalizumab. Second question on PD-L1, I think you started some trials in combination with ipilimumab, so which sort of patients PD-L1 lows or PD-L1 all-comers? And then on POPLAR and OAK, just wondering what sort of overall survival you've factored in from a Taxotere perspective there, whether it's sort of historic rates of five, seven months survival or 12 months as we've recently seen in (inaudible) data and if you could give us any insights on the powering of those two trials. Obviously, POPLAR is much smaller. And then finally, just on the wholesaler stocking that you mentioned affected September, just wondering if you could quantify that to give us some help on what might happen in October as well. Dan O'Day: In terms of lampalizumab, yes, our base case, our understanding is that one year with the geographic atrophy results, depending on the results, we feel that that could be a point where we could file at this stage. That's our base case assumption. And I understand your comparison to wet AMD. But that's so far the discussions we've had with the agency around dry AMD. By the way, I think it always depends on the magnitude of benefit versus safety profile that you see at the one-year. And so I'll just put that caveat in as well. On the PD-L1, I mean I understand your questions. I think both with the combination with ipi in bladder and also the underlying statistical and powering for OAK and POPLAR, I'm afraid those just aren't things we're going to disclose at this stage for competitive sensitive reasons. But I can assure you that we will continue to pursue combination therapies and that we really feel we have a very robust program in non-small-cell lung cancer between the four trials that we have accordingly. And then finally the wholesaler stocking in the United States, it's an impact, I would say, of around $50 million to $100 million in September, which we think we'll watch out in October. Remember as we change these distribution channels, we wanted to make sure that there were no stock-outs at the hospital levels. So we ramped, if you like, the inventory in the specialty distribution channel. We essentially kept the inventory in the general wholesaler distribution channel. And over the course of October, certainly into November, we'll see that come down again.

The last question for today is from Mr. Philippe Lanone, Natixis. Philippe Lanone - Natixis: One remaining on the HER2 franchise, because we have very strong US growth at Herceptin, about 10%, it was 17% last quarter, while Kadcyla is $70 million in absolute terms for the fifth consecutive quarter. So it seems that these figures don't fit quite well with the market share gains of Kadcyla in metastatic. And I wonder if there is any inventory situation that could explain for Herceptin. Dan O'Day: Kadcyla continues to be nicely penetrated in the second-line and beyond. I gave you the market shares before. And on the one hand, you think that would offset Herceptin use from that setting. I understand what you're saying, but we more than offset any cannibalization, I would say, of Herceptin in the second-line and third-line with the increased uses in the first-line. So remember, Perjeta plus Herceptin in the first-line has a very high penetration. And very importantly, the duration effect in the first-line setting is also longer. So between those two, it is really a true volume effect here, I'm going to have to say, that's driving the Herceptin sales growth in the first-line. And then the other thing is we continue to do well, of course, in the adjuvant setting. And in the adjuvant setting, we've now penetrated into the new adjuvant setting with Perjeta and Herceptin as well. So that's also increased Herceptin use overall. So it's a good news story on Herceptin, either being used alone in the adjuvant setting or the increased use with Perjeta in the neo-adjuvant in the front-line setting that is more than offsetting any Kadcyla penetration downstream.

This was the last question. Thank you very much for joining our briefing today. And I wish you a good afternoon and a good morning. Thank you for your interest in Roche. Bye-bye.

Ladies and gentlemen, the conference is now over. Thank you for choosing Chorus Call and thank you for participating in the conference. You may now disconnect your lines. Good bye.