Severin Schwan - CEO Alan Hippe - CFO Daniel O'Day - CEO Roche Pharmaceuticals Roland Diggelmann - CEO Roche Diagnostics

Tim Anderson - Bernstein Sachin Jain - Bank of America Richard Vosser - J. P. Morgan Sam Fazeli - Bloomberg Intelligence Vincent Meunier - Morgan Stanley Jack Scannell - UBS

Ladies and gentlemen, good afternoon. Welcome to the Roche Half Year Results 2017 Conference Call and Live Webcast. I am Sari, the Chorus Call operator. I would like to remind you that all participants will be 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 Mr. Severin Schwan, Chief Executive Officer of Roche Group. Please go ahead, sir.

Good morning. Good afternoon. Welcome to our half year briefing. Looking forward together with my colleagues to share a strong set of numbers for the first six months with you. Let me get right into the numbers here, seeing sales up by 5%, very much driven by our recently launched medicines and we see a strong development also on the bottom line with EPS growth, cash flow and IFRS net income accordingly. If we look into the divisional performance, both divisions up 5%. Pharma, very much driven by the new medicines, such as TECENTRIQ, OCREVUS, ALECENSA, also PERJETA doing very well as well as the immunology portfolio. Diagnostics, again as in the past, very much driven by the immunodiagnostics business, which continues to grow double digits. So overall, mid-single digit sales for the half year and it's very much in line with the development we have seen over the past quarters. Sales very much stems from the international region, China in particular as well as the United States and that again is very much related to the new launches, which we typically see first in the United States. So Pharma up by 8% over the first six months in the United States. Again, key drivers, more than half of the sales growth actually in the pharma division from the recent launches, OCREVUS really a phenomenal start, certainly even above our own expectations, ALECENSA doing very well. As you know, it is also approved for second line. We look forward to a pull in the first line setting based on the very strong ALEX results. And TECENTRIQ, on the market for bladder and lung as you know and very good to see that we now got also a positive CHMP recommendation for second line lung and bladder in Europe as well. As you know, we had major read-outs in the first half of this year. Let me just re-emphasize again the importance and the quality of the APHINITY data. You know, PERJETA today is already a medicine with sales of above 2 billion on an annualized basis. It has been growing with 17% for the first six months, so doing very, very well in the metastatic setting and now based on the APHINITY data, we bring this important medicine for the early stage setting. And what I'd like to re-emphasize is here we are talking about cure. We're talking about early stage breast cancer, where we have seen overall a 19% risk reduction of reoccurrence and death. So this is highly meaningful date, a few which is shared by many thought leaders in this area. And we do look forward to filing based on the APHINITY data and bringing this important medicine. Now also for the new indication to more women and I have no doubt whatsoever that this will be a substantial growth driver for us as we go forward. Operating margins, stable and particularly if you consider that we had this one-off PSI effect in the first half of 2016, also reflected in core EPS, which got an additional boost if you like from the financial income and lower interest rates. As the portfolio is concerned, it very much reflects our strategy, which is based on innovation, on breakthrough medicines. We stand now at 16 breakthrough therapy designations by the FDA. This is industry leading and it's this kind of innovation, which will enable us to offset the entry of biosimilars, which we expect increasingly so over the next years. Let's shift to the outlook for the current year. You all had the opportunity to see the data for Emicizumab in Berlin at ISTH Congress. So we have filed Emicizumab and hope to bring this medicine to patients as soon as possible in the inhibitor segment. There is other important readouts to be expected for the second half of this year with TECENTRIQ, Lampalizumab and HAVEN 3 for the non-inhibitor segment for Emicizumab. Just like to perhaps put in power 150 into perspective. I mean this is one of a number of pivotal trials reading out for TECENTRIQ over the next 12 years, sorry, over the next 12 months I should say. And TECENTRIQ is only one of 10 new molecular entities we have in clinical development in cancer immunotherapy. So this is a promising area for us. We invest accordingly in a very broad portfolio. Not all of the trials will be positive, but on balance, I think we are very much in a leading position here and we will develop this franchise as we go forward. Now, let me also comment on the outlook for the full year. You have seen we have raised the guidance to mid-single digit sales growth and I'd like to share some of the elements, which influenced the dynamics if you compare second half to the first half. So first and foremost, we expect a continued strong underlying sales momentum following the good launches we have seen in the first half. So we should see this to continue. On the other hand, we have as planned the first biosimilars on the market. And we have seen very limited effect for the first half of this year and this will gradually start in the second half of this year and then obviously continue into 2018. On the core EPS side, we had a negative effect in the first half of this year. You remember, we had the one-off income from the change independence scheme in Switzerland last year of over 400 million. But this was one-off in the first half 2016. So in the second half of 2017 this effect will washout. On the other hand we had more gains on product divestments from all the medicines in the first half this year. We expect this to be lower in the second half of this year. So this will be an offsetting effect for the full year. And I should also say that we had a significant income from the financials with the like of bond redemption on the one hand, some benefits from the sale of exited securities and again we would expect that to be lower in the second half of the year. Putting all of that together, we have raised the outlook now to mid-single digit sales growth, continued to expect EPS growth to be in line with sales growth and as communicated earlier expect the dividend to increase further in Swiss francs. And with this I'm happy to hand over to Dan. Daniel O'Day: Good afternoon, good morning from my side. Great privilege to present the pharmaceutical results for the first half of the year in 2017. Let me go right into the overall results, which was a 5% constant currency exchange growth and Swiss franc growth. And you can see the United States, really strong growth at 8%, driven by the launch of our new products there, Ocrevus, Alecensa, Tecentriq and the immunology franchise. Europe flat, good growth with Perjeta, with Actemra, some offset with Avastin, some of the Avastin - particularly in France delisting. But beyond France, the Avastin volumes and shares stay the same in Europe. Japan, we had the first quarter of back to continue price decline with Avastin from the previous year which is now washing out in the second quarter. And then international region of 5%, largely driven by Brazil and good strong growth from China as well. On the P&L, you can see the overall core operating profit growth at 3%. The underlying growth excluding PSI is around 7% and you see the line items here. We had some additional disposals in the first half of products, which caused an increase in the 19% on the royalty line. We don't expect to see those in the second half. Cost of sales really predominantly driven by volume and also some new manufacturing capacity coming online as planned and as expected for our new medicines. M&A very much in line given also the significant investments last year in the new launches and R&D roughly in line as well with the sales growth, with G&A growing at around 3% on the admin line. So this is the good news story, really terrific to see that the growth is driven by new products at exactly the time that we needed over the coming years. So Tecentriq, Ocrevus and Alecensa alone contributed CHF500 million to our growth in the first half of the year, more than half of the growth of the entire pharma division really coming at exactly the right time. The immunology franchise also very strong growth here driven by Xolair, by Actemra, Esbriet. I'll go into those products a bit. And you can see here Lucentis growing at 2%, we had a prefilled syringes launch in the first half of year, two new indications with CNV an diabetic retinopathy. And overall good stabilization of shares in AMD and in DME. Avastin was really a tale of two stories, we saw some pressure on Avastin to some degree in United States. We'll get to that with frontline lung. I already mentioned the breast cancer delisting in France, but good strong growth in Avastin in the emerging markets as we continue to make headway in the public sectors. Going to the oncology franchise, we have 4% growth overall for the oncology franchise. I'll cover Her2 and Avastin in the following slides. Maybe a couple of comments on the CD20 franchise where we had overall a 4% growth, but 3% growth with MabThera in oncology, 44% growth with Gazyva still in the relapse refractory setting. We will just beginning into the frontline follicular lymphoma setting in the autumn period in Europe and shortly after in the United States. It did receive NCCN guidelines in the United States for the frontline setting and we saw the EU positive opinion just recently in July, in Europe for the frontline follicular lymphoma setting for Gazyva. So again starting to get into the larger segments of Gazyva population in the future. We also by the way received approval and have launched now the subcutaneous formation of MabThera in the United States and Tarceva's expected increase competition. Tecentriq more than 200 million in sales in the first half of this year, driven by bladder and lung at about a 65% to 35% share of those two increasing share and lung, and positive CHMP opinion really terrific news for Tecentriq in Europe in both second line lung as well as first and second line bladder. And I'll get back to that in the second half of my presentation. Alecensa, strong growth, 103% driven in the second line setting in the United States and now Europe with their approval there. Strong growth in the frontline setting in Japan with a 65% market share and we look forward to bringing that frontline setting that has been recognized in NCCN guidelines to the US and Europe very shortly. Cotellic and Zelboraf, for growth Zelboraf sales has lost some sales and share but the combination therapy which is the new standard of care is gaining share and sales in the US and also in Europe. On the Her2 franchise, 7% growth for the quarter, Perjeta accelerating to 16% in the second quarter, strong demand really across all regions with Perjeta. At different stages of penetration into the metastatic and neoadjuvant setting but good strong growth for Perjeta. Herceptin in 4% predominantly volume driven of course, but also due to longer durations of treatment in combination with Perjeta. Some also you know penetration into the international markets as well. Kadcyla good growth at 7%, really across all the regions. We're really looking forward to the filing of affinity and bringing Perjeta into adjuvant setting. And Herceptin, we're seeing continued conversion for subcutaneous and Perjeta continued penetration into the first line and into the new adjuvant starting. With Avastin and overall flat growth for the second quarter. US, as I said before some competition in the lung cancer setting although [indiscernible] and we initially expected. EU, again outside of France very steady share development in EU. And international, a 15% growth, we recently received by the way in China the NRDL listing for Avastin, MabThera and Herceptin in China in a public setting. We will be working with the authorities there beyond what we've had basically, which is private pay and provincial reimbursement to now roll out our three medicines in a much broader way in China in the years to come. The outlook for 2017 two continued uptake in the ovarian cancer setting and of course we expect the first of many trials to read out. In the lung cancer setting for Tecentriq and of course we have the combination of Avastin plus Tecentriq in IMpower150 that will work out and the renal cancer with Tecentriq which also includes Avastin as well. Immunology, again really strong growth story with 8% for the quarter, 9% for the half year. Xolair at 13% continuing to grow in its indications of allergic asthma and chronic idiopathic urticaria as well. The pediatric indication drives growth not only in the pediatric setting but also in the broader franchise. Actemra, 12% growth, good strong growth in the rheumatoid arthritis settings, of course it's subcutaneous continue to drive the growth there as well as we've just recently launched the giant cell arteritis indication in the United States and received a positive CHMP in Europe. We look forward to the potential that those indications can bring to Actemra in the coming quarters as well. We also filed for an indication called CRS which is the utilization of Actemra in the current T-cell area in the United States and as I said, continued growth in the first line monotherapy leadership setting continued to increase our growth there as well. And MabThera in the immunology setting continued to have good growth in the indications of rheumatoid arthritis and vasculitis. Esbriet, we saw an uptick in growth from the first quarter, so 19% growth in the - sorry, in the second quarter now compared to around 13% in the first quarter. You can see Esbriet growing 20% in the United States, 13% in Europe, growth driven by the moderate segment predominantly, but also now more and more into the mild segment. The severe segment is fairly well penetrated at this stage, but the mild segment is only penetrated at around a 40% rate for any type of therapy. So there's still opportunity clearly in the mild segment and the moderate segment as well. We continue to have market leadership in the US and EU5 and we have programs now underway that encourage earlier stage adoption of therapy into the mild patient segment that are starting to show some effect also on our Esbriet sales and we'll continue to invest in that between now and the near term future. We were also able to launch the three tablet a day formulation in both the US and in Europe, so this will be a significant convenience benefit for patients as well. Clearly, I would say the highlight of the quarter was without a doubt OCREVUS, CHF190 million of sales, far beyond expectations - far beyond our expectations in the beginning when we launched it, driven by really good underlying demand across the indications, the feedback we're getting from patients and physicians continues to be very positive. The acceptance and the endorsement of payers in the United States also very positive. We have now, around at this stage and it's early days on our shared determination, around 60% of the patients that have been treated so far are with RMS and around 40% with PPMS. Clearly, there was more patients waiting in the PPMS segment to go on to therapy. So we would expect over time for that share to moderate and obviously to have more of our share move towards RMS versus PPMS. But I can only explain to you how strong the underlying demand here as we've treated around 5000 to 6000 patients so far in the first half of this year. We have many, many other patients that have been contacted through our Access Solutions, which is a surrogate for prescriptions, and of course, those patients are just beginning to convert now to their first infusion, which we expect to continue to have very strong trends in the coming quarters. The EU launch preparation is very much on track. We look forward to an approval there towards the end of the year and you may have seen another high incidence country, Australia, received a full approval in RMS and PPMS just this month. The strong, strong results on the sales side and in the product benefit side for patients driven by new products, let me just touch base on a couple of the really important read-outs that we had in the second quarter that we have presented in a meeting, so that I want to emphasize a bit starting with clearly one of the highlights, which was APHINITY in the adjuvant setting. I remind you we're talking about curative setting here, we're talking about around 330,000 women that are diagnosed with HER2 positive breast cancer a year. We also know today that as good of a therapy as Herceptin is, one out of four women over the course of ten years will go on to develop metastatic disease and developing metastatic disease certainly is something that shortens an individual's life. So this is extremely meaningful to look at the ability with PERJETA in combination with Herceptin already at the three year mark and close to four year mark seeing a 19% overall patient benefit and importantly in the more than 75% of patients treated with Herceptin today that are in the higher risk categories, that is the node positive hormone receptor groups, we saw a benefit of 23% and 24% respectively. As we talk to our thought leaders, as we consult with physicians that have a deep experience in adjuvant therapy and breast cancer in particular, we continue to hear from them how meaningful these results are for patients, how this type of a benefit in the past with other therapies has been very successful at shifting the standard of care. We've also had very good discussions with regulatory authorities. The filings are ongoing. We'll be alerting you as soon as those filings are accepted. But we remain very confident about the clinical meaningfulness of this and what this will mean for patients in the curative setting to avoid them, more and more patients from ever going on to develop metastatic disease. I would also say the subcutaneous co-formulation of Herceptin and PERJETA is something we have in early stage development and something I think that will continue to support us as we move forward. Moving forward to ALEX, the ALECENSA data at ASCO that was received with great enthusiasm, I just remind you of those results in the first line setting. This was almost immediately recognized in the NCCN guidelines as a priority. And you can see here the reason why. We went from against a very active medicine in this class survival on the median of less than one year to more than two years and the curves are still maturing. And extremely importantly I think is that the reduction in the CNS progression, very, very meaningful for these patients. It's four times as much in the comparator arm that it was with ALECENSA. You can see the results on the right hand side of the slide, which are quite stunning and quite promising for patients in terms of reducing the progression to the brain here. We have completed the filing in both the EU and the US and we're obviously working with the regulatory authorities to get this medicine to patients as quickly as possible. Already in Japan, where it has been approved for some months now, it has a share of around 65%. On the cancer immunotherapy strategy, we remain firmly committed around the science based approach toward cancer immunotherapy strategy. It really comes in three different waves. Wave one, it was getting to - getting TECENTRIQ to patients as a first indication in monotherapy, in lung and bladder cancer. Wave 2, step 2 is really around leading in combinations, in many disease states, including the frontline lung cancer setting, renal and also colorectal and in breast cancer. And then wave 3 is really around the opportunity to look at new novel enemies in our pipeline, combined with TECENTRIQ to get at those patients that are not being well served today by checkpoint inhibitors. And as you know, we have more than nine additional medicines behind TECENTRIQ in the clinic. We have another 10 that are pre-clinical that will continue to progress. Perhaps, some of the most exciting ones in our pipeline right now are the bispecific antibodies, both for hematology or the blood cancers and also for solid tumors. I've highlighted our solid tumor presentation from ASCO here, which is the bispecific antibody on CEA-TCB. With TECENTRIQ, in this case, in very difficult to treat third line colon cancer, a cancer that has shown not to be responsive to checkpoint inhibitors in the past and what you see here is a clear dose response, very encouraging early data. With this data, we're moving very quickly to take this into label enabling trials and you can see here on the right hand side that the cancers where CEA is over expressed, many of them are the cancers that we still have not been able to get to with checkpoint inhibitors alone. So this combination in early stages offers promise for patients and even in some of those cancers where we've seen some activity with a single checkpoint inhibitor, you also see that CEAs expressed there as well, such as lung and breast. I thought you might be interested in some of the news we had recently around IMvigor211. You have, I'm sure, seen that the EU regulatory authorities have recommended TECENTRIQ for approval for very broad indication in second and third line lung as well as both first and second line bladder and the second line bladder recommendation by CHMP was driven by the results in IMvigor210 and IMvigor211. And although we did not hit the primary endpoint in IMvigor211, the data was presented at the end of June at a conference and I just want to spend a few minutes explaining why we feel that when you look at the entirety of the data around TECENTRIQ that it's a very, very compelling story, a very consistent story around TECENTRIQ's activity in bladder cancer. And what you see here is the data that was presented on the left hand side of this slide, it's the all-comers overall survival curve and where you see the dotted lines is where the primary endpoint was identified and read out. And looking at this curves you can see that for a variety of reasons that we understand some and we're still trying to dig into. The curves actually separated quite late between the chemotherapy arm and between the Tecentriq lines. But they very clearly separate and the durability of response with Tecentriq is very compelling as you can see here. So even if you look at the 12-month point or you go out to the 24-month point, you can see the clear separation of the curves. And I think this is something of course that positions and patients and regulatory authorities are taking into account. To compare that, on the right hand side of the slide to the results that we saw from the Tecentriq 210 study which was the source of data used for the accelerated approval in the United States. You can see that the black and blue lines here almost mimic each other. In other words Tecentriq is behaving very consistently between these two studies. So what is behaving differently is the chemo arm. And as we look at the next line, you can see that when we break out and there were two primary key moments in this trial, one was the taxanes and the other one was vinflunine. And what you can see here is on the left hand side is just looking at taxane. So breaking out taxanes versus chemotherapy, you can see that in this post hoc analysis clearly, you can see and know that the significance is met when you look at the taxanes. And the vinflunine arm was the arm that performed better than historical standards at least in the early stages of the treatment of the disease. Having said that, even with the entirety of the data on chemos versus Tecentriq, the safety profile Tecentriq was significantly advantageous to that of chemotherapy and that's what you see on the right hand side here. So you can see that as you look deeper into this data it both validates the data that we've seen in the past in bladder cancer and in other cancers on Tecentriq and validates its use in the entirety of bladder cancer. So I thought I would give you a little bit more color on that as well both for bladder cancer but also for the entirety of the of the Tecentriq database. I had the pleasure of being at ISTH this year in Berlin and clearly the HAVEN 1 and HAVEN 2 results were met with great enthusiasm by the physician, by the patient, by the attendees at ISTH and this is the reason why. You can see here the results now in some detail that we published and also covered in the New England Journal of Medicine article. The reduction in bleeds, which is the primary endpoint for the study versus prior bypassing agents was 79%. And when you look at the number of treated bleeds, there are more than 70% of patients in the adult population that had zero bleeds at the time of the analysis. So really significant additional data for patients in the setting that have unmet medical needs. And with this data, the HAVEN 1 data we have proceeded with the filing in the US and in the EU as well. The HAVEN 2 study was also met with a great deal of enthusiasm. As one of the key themes of the meeting was with a disease like hemophilia that is constantly progressive and every bleed you have has an impact upon your long-term functionality but treating very early in the disease was a theme of the meeting. And so when we presented the HAVEN 2 interim results where you saw - of the eight valuable patients so far, we'll have many more patients, all of them having zero bleeds. This was really quite an enthusiastic response from the crowd at ISTH. And gives us a sense of urgency to bring this treatment option to patients in the inhibitor setting with both adults and the pediatrics. Obviously we look forward to the HAVEN 3 trial results in a non-inhibitor setting later this year. So with that I would just say that in addition to being a very strong financial quarter, a strong sales and profit quarter in the pharma side, it was also a very strong quarter for our pipeline. So I've spoken about many of these but the Alecensa approval in Europe, Ocrevus approval in the United States, Tecentriq approval in accelerated fashion in the firstline bladder setting in the United States and now recommendation for approval in Europe for the broad label across lung and bladder. The Actemra giant cell arteritis approvable in the United States and the recommendation in Europe and then [indiscernible] filing in the EU and US. So really strong performance of the pipeline at exactly the right time that we wanted and we needed four our business moving forward and a Phase III rates out I've talked about Perjeta and Alecensa and also the launches of the biosimilar, excuse me the subcutaneous formation of Rituxan in the United States. So with that I thank you very much for your attention and I will turn it over to Roland Diggelmann for diagnostics.

Good afternoon, good morning also from my side. It's a pleasure to be able to present the diagnostics results. And you can see the half year sales grew at 5% across the division, very much driven by centralized and point of care as well as tissue diagnostics. Diabetes care in a challenging environment minus four 4%, but then again looking at the totality of the business and in particular the clinical lab business, so centralized and point of care molecular diagnostics and tissue diagnostics growing at a combined 7%, which is a very good result. And this is also have been mirrored across the geographies, where you can see here is of course the emerging markets that grow at a faster rate as we would expect, growth in all of the regions. And then the E7 countries which we focus very much on growing at 16% combined here with standout are Turkey above 20%, Brazil above 20% and then as a very, very large and important market for us, China growing above 20% as well. I should also say that we see good growth, continued good performance in Europe and in North America excluding diabetes care. So the highlight, and I'll touch more on diabetes care here. You see the continues pressure on pricing in particular in the US on reimbursement across the world, while we also start to see the certain shift in high frequency test is primarily type 1 diabetes from the traditional strip-based testing to more of a continuous testing, the sensor-based technologies. At the same time, we also see a transformation happening in diabetes towards more of a holistic comprehensive approach, more patient taking ownership of diabetes management and in this light we have also made the acquisition of the company mySugr which is the leading digital platform, open platform for patient engagements, which already has more than 1 million patients that really are able to share, track their results and manage diabetes in a more comprehensive way. Then molecular diagnostic is up 1%, virology negative by 1% here what we have seen is certainly a bonus of HCV testing as part of a curative drug the acting antivirals for hepatitis C. We've on the other hand seen a continues big growth in HIV and as you can see there is well blood screening up 3%. On the P&L, the 5% percent on the sales line that translate to 5% core operating profit. A few points worth mentioning on the royalties, extraordinary income on licenses in the first half that won't repeat in the second half. Then cost of sales at 4%. M&D at 5%, pretty much in line. R&D comparing to a strong R&D in the first half of 2016, so you should expect that R&D expense to grow again in the second half of the year. And then G&A, Admin plus 2%, very much in line as well. Shifting more to the products and the portfolio where you can see that immunoassays continue to contribute strongly, we've now had 19 years in a row with a double-digit growth, pretty much across the broad spectrum of immunoassays and there's a couple of the diseases, test areas on the right side. And it's really about the most comprehensive menu, a very broad menu. And then couple as you can see on the left side of the screen, a very strongly automated workflow approach toward the efficiencies in the lab and this is really what makes the strength of this franchise, on one hand high testing efficiency coupled with broad menu and a high medical value as an offering. And it is also very much what we have continued to execute against, notably the US launch of the cobas e 801 which is high throughput immunoassay module as part of the very throughput cobas 8000 family for the large laboratories launched in Europe and outside of the US last year was a very, very good reception in the market and we're in the process of launching now in the US and this comes very timely with then also the HIV approval by the FDA, which was the biggest gap that we had in the US menu, which is now closed, which also will allow us to continue to grow in the US immunoassay business. Shifting from centralized to point to care, and here, what you can see is the Liat, the desktop system for true PCR results, typically in about 20 minutes. We have focused a lot on the respiratory panels for the US market. You can see the panel there was influenza A/B, Strep A and others. And now shifting more towards the European, Asian settings where we're focusing on a menu for hospital acquired infections, MRSA and Cdiff initially and we'll continue to evolve with this menu also across more of the Infectious Diseases. Shifting to tissue diagnostics, which grew at 13% for the first half year and here too, a very good performance and also a focus on high automation. Looking at the entire work flow from the patients' specimen, all the way to the actual diagnosis and the opportunity to provide integrated solutions here. What you see here is an automated multiplex immunohistochemistry staining combined with an automated whole slide image analysis and that really allows for deep reviews through, for instance, spatial characterization of tumor micro-environment, but then also the interpretation of tumor infiltrating lymphocytes. So an end-to-end approach. We'll continue to develop this as a more of a disease specific diagnostics and then of course combine this with our PCR approach, with our sequencing approach to really provide comprehensive solutions to laboratories. And then a bit more of a look towards the potential future of diagnostics and here what's very exciting is the opportunities in circulating tumor DNA and the measurement of it in plasma. We have now launched three fully kitted options for ctDNA, one with 17, one with 77 genes, more in the tumor profiling area and an expanding tumor profiling and then with 197 genes as a surveillance kit. So, these reagents and kits include full informatics for sampling result, out result approach, excuse me. So really again looking at the sampling, all the way workflow to the result and these are the first, which actually are able to cover the four mutation classes in one panel. Very excited about the prospect here of continuing to having the opportunity to provide ctDNA as opposed to or moving liquid biopsy as opposed to tissue biopsy with all the opportunities, for instance, in patient monitoring and disease and management in a more comprehensive way. And then moving to the key launch list for 2017 and you can see that it's been a good half year already in terms of the portfolio evolution. I won't repeat the launches that we've had and then we have coagulation, our own solutions for coagulation testing coming towards the end of the year and also the CCM High Speed, which is the workflow automation for high speed laboratories also coming later this year. And with that, thank you for your attention and I will hand over to Alan Hippe for finances.

So thanks, Roland for the handover. Some comments here on the financials, I'm happy to wrap that up. Evidently, it has been quite a successful half year and here, you see the highlights. And as said already, strong operating performance on the business side allowed us to come up with a significant cash flow and you might have seen it already, the operating free cash flow has really grown by 2 billion compared to last year, which then allowed us to bring that debt down. Already mentioned, the good net financial result and I will shed some light on this one later on and as well as the IFRS net income, which has grown by 2% and here the growth opportunity was a little bit mitigated by some impairments that we have taken. And when you look really at the comprehensive set of our key figures and I will go through them in my presentation, you see really sales grown by 5%; core EPS up with 6% and you see the operating free cash flow up significantly and you also see that the free cash flow even has a higher momentum. Let me talk quickly about the free cash flow as this is a figure, which I won't dig deeper into in my presentation. This was supported certainly by the operating free cash flow performance and then in addition, we had lower outflow for interest expenses, I think something which relates very well with the financial results. We had lower cash out for equity transactions. We did Flatiron last year as you might remember and the last one was 100 million, that's cash outflow for taxes. And the other point to mention here on this slide is the currency deviation here. When you look really at the reported numbers in Swiss francs as well as in constant rates, I think a very small deviation here. So, I think it will surely change soon. Group operating performance, look at the P&L. I think my colleagues did a great job here, really leading you through this already. Royalties and other operating income has definitely boosted by divestment gains and the difference to last year is 181 million and this increase certainly is not reoccurring in the second half. Cost of sales, I would say that's volume and our rebuilding line with the sales growth and you look at M&D, very disciplined, given all the launches that have been described already and then R&D I think, really which goes well with sales. Then, you see G&A and a significant increase in G&A, which is certainly triggered by the positive effect of plus CHF426 million last year first half for the POS service income effect, admin overall grew by 3% and also shows a high level of discipline. Core operating profit, up 3% and as said, a little bit mitigated here the growth by this positive effect that we have seen in the first half of last year. When you look at the margins, don't want to dig into this, I think a nice stable development as you can see in all divisions and in the group as well. And that leads me to the core net financial result. Here, a couple of positives and let me start really with the point where we have improved that by CHF262 million. Well, one major boost was we didn't do a debt reduction yeah, as we did in the first half of 2016. Interest expenses came down and that's really an effect of our constant reshuffling of our debt. We have brought the interest cost down from around higher than 5% in 2012 to now below 3%. Equity securities, I think we had some transactions in equity securities, gave us a positive 53 million, also something which won't reoccur to that extent in the second half and that is underlining the positive development in that area. Well, the group core tax rate, nothing has really moved here. When you look at the effective tax rates, so on page 28 of the finance report, you will notice that the tax rate comes a little bit down, which is basically driven by the reversal of contingent considerations. Let me lead you from the core results to the IFRS net income, explaining with that, the non-court items. And the core operating profit, as said, has grown with 3%. We then had a couple of restructuring plans, nothing new, pretty stable compared to the first half 2016. We have the amortization of intangible assets, it was 906 million in the first half of 2017 and [indiscernible] is 583 million of this 906 million. Then, we have impairments of intangible assets and that's a string of impairments that we've done. There is [indiscernible] there is also Esbriet, which accounts for the vast majority of that impact and then you have the alliances and the business combinations, was a positive number, but there is nothing else and the reversal of the contingent considerations that I've mentioned already was a plus 197. Legal and environmental, we have released a couple of provisions and that leads us to increasing IFRS net income and increases 2%. Cash flow, quite a story in the first half 2017 and that also applies to all divisions and then to the group in total, strong performance. Let me dig into this. Operating free cash flow went up by 2 billion and what you're seeing is that the operating profit, net of cash adjustments, went up 844 million. So really that's the - the reason here is the underlying operational performance, which is certainly great. And then you see the networking capital movement that we have here, an improvement of over 1 billion. Networking capital has still increased in the first half of 2017, but to a much lower extent than the first half of 2016. So quite nicely and gives us a nice boost when it comes to the debt development. And for me on that slide, I think the story is on the right hand side because we have brought net debt down from the first half 2016 to the first half of 2017 by roughly 4 billion. So we are now at 14.2 billion and that's only driven by the strong operating free cash flow, which I've explained already. When you look really at the year-end of 2016, we were at 13.2 billion, now we are at 14.2 billion, basically what drives our net debt development in the first half it's a dividend payment. And as you can see, we have basically compensated that already. And you know we want to get into a range for net debt to total assets between 0 and 15%. We're now at 20% and we have reached that ratio by 5 percentage points within one year and it appears like that we can be rather optimistic to get into that range between 0 and 15% at yearend.. A quick look at the balance sheet. And one general development that you're seeing is that assets come down as well as liabilities come down. Let me lead you through this quickly and let me emphasize this is now a comparison from half year to end of the year 2016, so when you look at cash and marketable securities, reduction is basically driven by the dividend payment. Other current asset is pretty stable. And then you see the non-current assets, which have reduced basically based on the impairments that we have taken in the first half. Current liabilities, we have paid down some short-term debt. The non-current liabilities, these are the pension obligations going down based on higher interest rates and then you look at the equity and we were able to increase the equity ratio to 35% despite the fact that we have paid a dividend in the first half of roughly 7 billion. Currency, basically a wash that you can see, you see really on the left hand side the growth in constant rats, you see on the right hand side the growth in Swiss francs basically 5%. Swiss franc has weakened against the US dollar and some Latin American currencies and has strengthened against the euro as well as the British pound. And then you have our famous projection for year-end which is nothing else than modeling as we assume that the currency rate at the end of half year 2017 remains stable until the end of the year, which is not very likely I have to say. But you see really at half year basically or no impact if all the currency rates remain stable, we still expect rather low impact at year end basically on all lines. And with that once again the outlook which is raised for group sales, now mid-single digit for core EPS growth broadly in line with sales growth and for the dividend further increase in Swiss francs. And with that we are happy to take your question.

[Operator Instructions] The first question is from Tim Anderson from Bernstein. Please go ahead.

A couple of questions. Just your view on Astra's Mystic results today, I'm guessing you view this as clear validation for not having [indiscernible] and also validation for at least initially going against the grain with your chemo combo approach. So, just any high level perspective you have on that and whether you think this is a meaningful piece of the puzzle that has fallen into place or do you think we still need to see various readouts going forward to kind of think about longer term market shares across the different players? Second question is on APHINITY, really just from a regulatory perspective and markets like US, Europe, Japan, could the uncertain markets like Europe for example that you don't get adjuvant approved or that it gets approved but in a narrower setting than the entire population that you studied. So it's not a commercial question, just a regulatory question. Anything to worry about here or are you not worried about regulatory and it all comes down to commercial execution.

Thank you, Tim, Dan you want to comment on [indiscernible]? Daniel O'Day: Absolutely. So I think first and foremost, I would say that our plans remain exactly as we have them, which is to focus initially on the chemo combo in the frontline setting with and without Avastin. So I think nothing has changed relative to our plans as a result of Mystic. I would say that when you get the topline data report, you always have to see more data. So I think we need to see more data. I would say that in general the hypothesis that Mystic was going after was very different than the hypothesis that we were going after, right. I mean we are adding a single agent first and foremost Tecentriq to a known beneficiary if you like in the frontline lung cancer setting and in a lot of patients chemotherapy. And then of course the IMpower150 we're doing that with and without Avastin based upon some of the data that we've seen also in our renal trial, but still have to see whether that plays out in lung. I would say that the Mystic trial is just on the surface, is a trial that's testing a very different hypothesis, right. It's testing whether two checkpoint inhibitors, where neither one has necessarily demonstrated a particular effect in the frontline setting, can benefit either alone or combined against chemotherapy. So very different approach, we took an approach that we thought was right and we certainly still believe very much in our approach of chemo. I think both from our own data also seen the Keytruda data which is early data as well. So we very much look forward to the entirety of the readouts on our lung cancer program and I would just remind you it's a pretty comprehensive program. I mean overall we have seven trials, five of those are going to readout between now and this time next year. Of course our co-primary endpoint, PFS and OS, we remain to see whether both of those endpoints readout at the same time or whether there's an initial readout and the other end point comes, we have an ability to file based upon positivity on either end point, but which gives us a great deal of flexibility. In addition to our non-squamous readout with and without Avastin later this year, the IMpower150. I remind you that in squamous and also in the small cell which will readout in the first half of next year, we're very competitive in terms of timeframes vis-à-vis the competition. So have the ability to readout we think first in some of those cancer settings and lung cancer. So I would just say you know probably as much as I can comment on a study that we've just seen the topline results on. It certainly doesn't do anything to dissuade our strategy, it only reaffirms our strategy and we stand firm on the entirety of our program in lung and basically beyond lung. By the way in addition to the five trials readout on lung cancer between now and next year, we have three additional trials. So we've got the renal trial, we've got the triple negative breast and we also have the colorectal cancer. So as I've always said, in cancer immunotherapy we're still learning a great deal. I like our program and I think next year this time we're all going to be a lot smarter about what works in cancer immunotherapy and what doesn't including those agents, those new enemies that we're working on that have new mechanisms that we think are going to be able to attack things that the checkpoint inhibitors can't address alone or can't address with chemo combos. Relative to APHINITY, I'll just be very direct, I mean the conversation that we're having with regulators are all very encouraging very positive. As I said I don't see any risk of not getting approval for these adjuvant setting in Europe and United States as I sit here today. Whether that will be for ITT population or will that be approval that may be focused a bit more on high risk patients is a discussion I'm sure we'll have with regulators as we go through the process. But nowhere on these discussions do we see hesitation by the authorities to not see the value this brings to the frontline cancer setting. So I hope that answers your questions.

Can we have the next question please?

The next question is from Sachin Jain, Bank of America. Please go ahead.

Sachin Jain, Bank of America, two questions on similar topics if I may. So firstly on APHINITY, you've referenced regulator discussions, I wonder where you sit on cost benefit analysis and where you could train existing projected price or the benefit and how hat stacks up versus your existing oncology drugs and whether you've had any early peer discussions in the US and ex-US. And the second just a follow-up on a comment you made on the last question, would you just remind us of the timing for IMpower150 of PFS versus OS and whether they come concurrently or separately? And then a related question, you've, in a few comments, both Severin and yourself have made comments on the entirety of the lung program. I just wanted to understand that, it's just that just focus on the fact that we shouldn't focus on one city, was there any difference in relative confidence of IMpower150, which is the first dataset versus for example 110 and 130 where you have a limited [indiscernible] which are closer to where you have had stronger data and where Merck have had stronger data? Daniel O'Day: Thanks, Sachin. On the APHINITY, on the cost benefit side, I mean a couple of points here. Again, as you know when you go to your pharmacokinetic models and you start to run ICERs and other pharmacokinetic mechanisms, you have very different parameters in the adjuvant or curative setting than you do in the palliative setting or the late stage setting where you're extending life by months, fortunately with some of our medicine, sometimes years, but curative settings have a very different economic threshold if you like. And just to repeat what we said at ASCO, I mean when we look at the ICER models, even with the current costs of PERJETA and Herceptin in a market like the US or in other market, we get very positive readouts on those ICER models based upon the entirety of the ITT read-out and certainly in the node positive hormone receptor negative subpopulations, which again is 75% of the entirety. And then on top of that, of course, so that's statement number one. Statement number two is that over the time period that APHINITY is both registered and goes before payer authorities, we would expect first in Europe of course to have biosimilars become more prevalent and therefore the price level of Herceptin as a part of the combination to be reduced, which makes the economic - economics of the combination even more favorable, even in those areas where we're unable to do combination based pricing, which is obviously what we've done to date and even where we've done it today, there is a metastatic setting. We've been conscious to preserve the PERJETA price and to give, if you like in the Herceptin price. So I have a say our analysis and our assessment is that given the strength of the data and given the fact that this is a curative setting that this has a positive cost benefit ratio. On the IMpower150, well maybe, let me answer your third question and then go to the IMpower150. The reason that Severin and I are emphasizing the entirety of the program is to make sure that you understand the breadth of our program, because it's - we made a very significant move a couple of years ago when we saw this and we were one of the first to see I think the chemo TECENTRIQ combination and we said there's something here. We believe there's something here, we want to go very broad in this setting, squamous, non-squamous, small-cell, with and without different chemo backbones and with and without Avastin. I still believe that's the right move. There's - everything we're seeing the Phase 1 follow-up continues to support that or the competitor data. So the reason we're emphasizing that is I can't tell you that every single one of those trials is going to read out positive, whether readout positive on PFS or OS. I mean, science has to speak. It's no lack of confidence in any single trial, but clearly the combination approach, the number of trials we have given us even greater confidence to be able to change the standard of care in different aspects of the first line lung cancer setting. So I wouldn't read anything more into it than that. On the IMpower150, the PFS, OS. As you know, it's a co-primary endpoint. We expect it to read it. At the same time, we've been telling you it's probably quarter four, quarter three, quarter four, but it's event driven and it's, I would say, more likely in quarter four at this stage. And when you do the data cut, only then will we really be able to determine whether OS is mature, PFS is mature and how that will read out. So there are several options, at least two different options, one is if PFS reads out and OS is not mature, and then we carry the trial into OS to be mature, which OS is the gold standard or it's possible that it reads out and PFS and OS are mature at the same time. So I can't tell you that today. I know we've powered the studies well. It is a first trial read out for IMpower150 in quarter 4 and the trial will continue until all the end points read out, which may go into the beginning of next year at this stage. So I hope that gets at your question.

The next question is from [indiscernible].

Good afternoon and thank you for taking my questions. Dan, I would be very keen to hear your views on how the OCREVUS launch is going. Obviously, heard your comments, but it would be great to get more insights into how much of the 192 million we're seeing in the second quarter is likely to be stocking, how are you seeing that split between kind of RRMS, PPMS, where are you getting those patients from? And then secondly moving on to emicizumab or ACE 910, just help, given kind of the differential pricing for the two markers that emicizumab could potentially be used in, any thoughts on how we should think about how you're going to commercially position emicizumab. Thank you. Daniel O'Day: So yeah, OCREVUS, let me try to give you a little deeper color to it, again just my congratulations to the team and to the good launch in the United States, the Genentech colleagues and the entire team behind this. I think it really validates what we've been saying is that the MS market is looking for highly efficacious medicine with a reasonable safety profile and a very convenient dosing. So that's what I hear when I go around and talk to physicians and patients. And as you know we've presented a number of follow-up data points around the continued disability benefit that OCREVUS has vis-à-vis the comparators and I think that's important. Now, maybe the CHF192 million is really demand. It's absolute demand. We don't have major stocking in the channels. It's demand driven and it represents, as you know and you can do the math, the first of the infusions, many of the patients that were infused before the end of June will get their second infusion of course before the end of the year and then we continue to have the ramp and run rate on new patients coming into the mix. We still have quite a backlog as I said of patients on both RMS and PPMS, prescriptions we know have been written and they're waiting to line up infusion time and then every day, we get new prescriptions from patients in both RMS and PPMS. As I said, I gave you roughly and it's rough, the 60-40 split between RMS and PPMS. Clearly, we feel there's probably more PPMS patients that were waiting for the medicine than RMS, but there's a certain bolus on both indications. There is no precedent for this right in the MS field. All other indications have launched only into the RMS field, so you didn't necessarily have - to have the concept of having PPMS bolus versus RMS bolus and trying to figure out if those two are different. I think we're still learning about that, but I clearly expect the 60-40 to increase in terms of RMS patients over time, just in terms of pure population dynamics. And then maybe one other piece of color on the RMS side, I mean very encouragingly, it's across the spectrum of severity of diseases. So we have of course quite a few patients that are later stage that are switching for medicines like TYSABRI. We also have patients in the middle, switching from medicines like TECFIDERA, the interferons, but then we also have naïve patients as well and we're still getting a handle on that market research and that data, but I think it's fair to say that there's a good distribution, I would say, across the different levels and different cares of treatment there. And then finally at emicizumab pricing, I mean it's just too early to comment on emicizumab pricing, but you rightly pointed out, I mean, we expect and we intend to go into presuming successful HAVEN 3, the non-inhibitor and the inhibitor population and therefore we want to make sure we price the medicine and have pricing programs and access programs and that make sure the medicine gets in the hands of patients that need it. That's all I would say at this stage and when we have the approval, we'll certainly let you know the pricing policy.

The next question is from Richard Vosser, J. P. Morgan.

A question back on APHINITY, just you referenced discussions with KOLs post the data, just when you're discussing with them, just wondering how they're thinking about treating patients post the data. Are they looking to treat with both neoadjuvant and then go onto adjuvant treatment, so 15 months total or are they just looking to treat adjuvant going forward. So the total treatment for a patient will be 12 months? So just some color there would be great. And then going back to IMpower150, you haven't released data around the combination with Avastin and Tecentriq and chemotherapy. So just wondering how representative that data the Keytruda data of that combination with the experiences of the combination in your hands both in terms of efficacy and in terms of safety. And then just one follow up on the Mystic. Just wondered on your perspectives PD-1 versus PD-L1 and whether you see them as having equal efficacy on both PFS and OS. I recall Oak obviously having very good comparable data on OS with [indiscernible] and Keytruda, but I think the Mystic result today has thrown up for some people on their monotherapy result that maybe PFS or PD-L1 is a little bit different. So just some perspectives there would be great. Thank you very much.

Thank you very much Richard. Sorry, I have to go back to the first one. How the KOLs maybe treating in the world of Perjeta with APHINITY. I think clearly as you know, I mean we're making significant inroads into the new adjuvant treatment population around the world, still seeing growth there. I mean obviously that's one of the key drivers of growth that you see in the figures. In the first half of the year, the 60% Perjeta. And I would say that, I don't think there's a particular common answer today amongst KOLs exactly how they would go about training if they treat pre-surgically and how they treat post-surgically. I think in general the APHINITY data supports and have supported not just the APHINITY data but other data, a 12-month duration of treatment with adjuvant therapy whether that done pre-surgical - combination of pre-surgical and post-surgical or just post-surgical. So I think there will be different treatment practices there. I think we'll have to see how KOLs experienced that. But I do believe that the data suggests that 12 months is the right duration of treatment and whether that's a combination neoadjuvant or pure adjuvant I think we'll see as we go. And obviously we'll continue to talk with top leaders as we move. On the IMpower150, again I think we're not - I mentioned the Keytruda data as if you like supplementary data that continues to give additional data on the evidence associated with the chemo combo. But I would say more precise and more telling is our - it's that data in our own hands. So the Phase 1 extension data that we've seen with a variety of different chemo backgrounds granted. I mean when you get into different chemo subsets you get down into smaller figures and smaller numbers, which is exactly why by the way we designed such a comprehensive late stage program, it's hard to tease those chemos apart and to determine whether or not some of the early signals we saw in our Phase 1 may be indicative of what we see when we get into large numbers and true statically significance. So I think we will see how that plays out. But again, the evidence continues to point at this PD-L1 effect on top of chemotherapy being beneficial. So we'll see how that differs between chemo back. And again we'll have other competitive data reading out of course this fall as well. Remember the Keytruda data was based upon overall response in PFS. So we have yet to see kind of a more comprehensive dataset from Keytruda, which we'll also see later this year as well. And then finally on the Mystic, I think again those are all good questions to ask Astra I think. I don't know and we don't know how the study was powered versus the different subsets, the diagnostic lean rich and non-diagnostic lean rich for PFS. I think that'll all be interesting information when the data is published. So I would steer clear of that. And on your broader question of, do we yet have enough evidence to suggest that there's a difference in PD-1, PD-LI, I would say that evidence is all still developing. And again we'll be a lot smarter in a year's time. We may be able to definitively say there's a difference or definitely say there's no difference, but to date I think we've got contributory data points, we've got conflicting data points. So let's see the entirety of the data and see where it goes I think for us, we have no reason to believe that PD-L1 is certainly not an inferior way to attack it and may or may not have some advantages, but beyond that I think again our thoughtful approach to designing our programs, designing the clinical trials and determining what's indications we can take leadership positions in and what we may not take leadership positions in both in terms of where we stand in our programs and how we can progress that's really what we're focused on today.

The next question is from Luisa Hector from Exane. Please go ahead.

May I ask the operator to switch to the other set of question please? This seems to be the wrong conference here.

The next question is from Matthew Weston, Credit Suisse. Please go ahead.

That sounds much better, thank you. Please go ahead.

Mr. Weston, your line is open. Maybe your line is muted.

Should we take the next question instead?

The next question is from Sam Fazeli, Bloomberg Intelligence. Please go ahead.

Please excuse me for going about 150 one more time, it just we've just come out of Mystic reporting this morning and two months ago the CEO gave a perhaps a less than glowing expectation of Mystic saying, it's not the only thing we have, it's just the comments that they made earlier on just made us, but made me certainly think a little bit along those lines. Just to give us some more comfort on that. And also on 150, which arm will you be testing first, the Avastin plus Tecentriq plus chemo or the Tecentriq plus chemo as the primary endpoint readout. And then lastly on bladder cancer Phase 3, when we look at the curves, the OS curves for keynote 45 verses IMvigor211, it does look to be similarity, although I know you've said different, similarity in the chemo arm, if anything, the chemo arm in 211 performed perhaps a little worse in the longer term. But it's the - it's treated arm, the PD-1, PD-L1 arm that looks a little bit different there. Should we - going back to the PD-1 versus PD-L1, is there anything to think about there or are you going to suggest that these are two different trials and we should think different, just not look at them like that?

Let me turn comment on your overall question around cancer immunotherapies. I mean, we are deeply convinced that cancer immunotherapies offer a great deal of additional medical benefit to patients. And this is why we have reallocated substantial resources exactly into this field. So overall, we are very bullish about this segment and we want to take a leading role and what I emphasize was the breath of our portfolio. And for Tecentriq alone we have eight studies reading out over the next 12 months not only in lung cancer, many in lung but also in other areas such as breast, renal for example. And on top of it, we have ten new medicines in clinical development. I just believe that with a broader portfolio, many shots on goal here. Multitude of potential combinations, we have better chances to be at the lead of a very promising field. That's all you should read into it and we remain very, very bulling about TECENTRIQ. We believe this will be a huge growth opportunity for us and I just remind you again that we just got positive CHMP opinion for both lung and bladder cancer. So, we are off to a good start here and let's see how the data develop. Dan? Daniel O'Day: Yeah. Thanks, Severin. I couldn't agree more with Severin's statement. I mean, it's very difficult to compare across companies. I think one of the reasons we're emphasizing the entirety of our portfolio is because we have a very broad portfolio in lung and that we've been in this business long enough, you can't interpret any one trial. So I wouldn't take it as a lack of confidence, I wouldn't take it as definitive that all trials will be positive. They won't. So we're just trying to put it into context, but our concept around IMpower150, I'll say this and the reason that we had advanced this study, nothing has changed, nothing has changed. We still believe in the chemo background. We still believe that Avastin could have a synergistic effect and I would just say that, to your point around what will read out? Well, the trial will readout - all three arms of the trial will read out. So you will get information on the first read out about how the, both the, if you like the TECENTRIQ plus chemo arm does and the TECENTRIQ plus chemo plus Avastin. So you'll get a view on all three arms and we'll see what we see in the data obviously. So, yeah, so I just want to make sure that you understand why we're putting our studies in to context, because we have a lot of studies and we want you to be aware of them over the course of the next 12 months. And then finally on the IMvigor211, yeah, I guess I see it - I'm not sure I see it the same way you do, Sam. I mean the active arms look quite similar. I don't see a discrimination with all the caveats around, this is cross trial comparison, right and population base and different dynamics. But if I put all those caveats around it, the way I look at the data is the active arm looks similar and the chemo arms look different. So I may be looking at it differently than you do, but that's the way I see it. But I would even shy away from that and just get back to the 211, 210 durability of response. I think the durability of response is maybe something I didn't mention before, which is significant. Just to remind myself, I mean it's, in the IMvigor211 trial, we have 21 months of durability with TECENTRIQ and 7.4 months of chemo so far. And I think that is, yeah, forget about cross trial comparisons. That to me suggests an effect. And again at least one regulatory body has seen it that way as well.

Next question is from Vincent Meunier, Morgan Stanley.

The first one is on HAVEN 3. What level of ABR you would consider clinically meaningful to drive a substantial use in the non-inhibitor setting. Could you, for instance, target something like two or below two? So one question on PERJETA adjuvant and the subcu formulation, Herceptin plus PERJETA. Can you tell us what you need to do and what is the timing for that project? You, I mean, and the linked question is, how long should we have to wait for the launch of the combination and what will be the time difference between that and the approval in adjuvant. And last question on OCREVUS, can you explain a bit more in detail of the pattern, the expected patterns of the sales ramp up, given the cell-dosing regimen, which is quite specific and how do your charge patients, do you charge them on a six-month basis or on a monthly basis? Thank you.

Dan, over to you. Daniel O'Day: I'm afraid I can't give you a whole lot of additional color, predominately for competitive reasons on HAVEN 3. We haven't disclosed exactly what the ABR difference is. I mean suffice it to say that the, we're very encouraged by the phase 1 trial in the non-inhibitor. We recruited the trial very fast. Good anecdotal reports but that we have to wait to see how the trial reads out and we remain confident, as confident one can be prior to the Phase 3 trial read out that what we've seen prior to this in terms of the data and the difference is meaningful. And we'll see how that reads out. On the co-formulation, PERJETA and Herceptin, as I said, we're still in the early stages of that development program. We wouldn't expect that until around 20 or 21. But I remind you that we've got obviously PERJETA in the entirety of its programs, rolling out in the United States and elsewhere. We've got Herceptin subcu which is up now to around 50% in the launched countries. And we'll be looking at Herceptin subcu after the success from MabThera subcu as well in the United States. So I think we've got a variety of programs that allow us to continue to compete, if you, like against the biosimilars, the most important was in changing the standard of care, but the convenience associated with subcus is substantial. It really is substantial and we think that will give us a benefit between now and when you may have some co-formulations. And then finally on OCREVUS, as I said, I don't know if I can give you a whole lot more in the pattern. I mean I think in general, the payment is done every six months. It's not done every month. Certainly the way that we're accruing for the payments, I mean with the way the drug is purchased I should say, not the way - the way the drug is purchased is purchased as a dose. And then the dose is given to the patient and that's given at the time of the infusion. So the 190, if you like, is reflective of - I mean you can take the number, you can do the math. If you take CHF190 million, you need to divide it by around 5500 patients. You get the, that's at around $65,000 a year. You get this, the numbers work out. So it's that, you can see that we're selling into the market by infusion if you like. And the pattern of the ramp up again, I mean you only fully know what the - whatever bolus effect could possibly be when you're looking in the rearview mirror. All I can tell you is that I - when I take a look at the number of calls into our access lines versus the number of patients that have been dosed, we still have significant demand that continues to need to be fulfilled over the course of the next coming quarters and we still have very robust demand on a daily basis. So every quarter, I'll do my best to put more color on that. I'm really giving you as much color as I can right now, given the fact that the date is imprecise that's coming in to me. But as the data come in, I'll let you know. This is definitely a very strong launch and it's definitely continual demand driven as I look at the underlying basis of the business.

Dan, thank you very much. We are approaching the end of our call. Can we have one more question please?

The next question is from Jack Scannell, UBS. Please go ahead.

Hi. So you've - just one question, you've raised guidance and that may be because some things are going better than you expected such as the OCREVUS launch, but also it's possible that it means that some things that could have gone badly haven't. Specifically, can I ask you if you've got any insight into sort of biosimilar attraction in Europe in the second half of the year.

Well, by and large, biosimilars enter the market as we expected. So we are now seeing the first biosimilars for MabThera. We expect biosimilars for Herceptin only towards the very end of this year, early of 2018. So really the guidance is driven by the very good launch of the new medicines, which are at the very high end of our expectations. So this is, ALECENSA doing very well. This is good continued growth for TECENTRIQ and of course very much so, the very successful start of OCREVUS. So this is really driving the sales and that has been the reason to upgrade the guidance. Thank you very much for your interest and have a good day. Thanks for joining.

Ladies and gentleman 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.