Good morning. Good afternoon everybody. Thank you so much for joining us today. I'm Pascal Soriot. I'm the CEO of AstraZeneca. Welcome to the full year and the Q4 2015 results presentation for investors and analysts. We are here live in London, and I know there's quite a number of you on the telephone, on the Webcast. There is a Webcast on astrazeneca.com, and the presentation is actually posted online for those who want to download it. I'm joined today by Luke Miels who is here, our Executive Vice President for Global Portfolio and Product Strategy, Global Medical Affairs and Corporate Affairs. Sean Bohen is our EVP for Global Medicines Development and is our Chief Medical Officer; and here also is with me Marc Dunoyer, our CFO. In addition we have Mondher Mahjoubi in the room, Head of Oncology and Global Portfolio Strategy, who can also help us answer some of the questions in oncology you may have. We also have a number of key members of the AstraZeneca team here from Iran [ph] and some others as well. It's really great to see so many of you here today, despite a very busy reporting season. I'm sure that you've been extremely busy in the last few days. We look forward to taking you through the results and our achievements in 2015. If you may -- if I may ask you, if you could turn to Slide 2, please, this is our forward-looking statement. And then please turn to Slide 3. The plan today is for me to provide a short introduction, and then I'll hand it over to Luke who will give you an update on our Growth Platforms and the launch of New Oncology. It's a new Growth Platform, the number six. He will define what we call…

Thanks, Pascal. So it's a pleasure to present our product sales results today. Please turn to slide 8. I'm going to spend a few minutes just talking through the Growth Platforms, which grew across all areas. This is encouraging, naturally, because it's on the back of strong performance across these platforms, and these platforms are critical to our long-term goals. I'll review each of them individually and in more depth later, but broadly Respiratory was driven by strength in emerging markets and by products in the US and the EU. Brilinta continued to enjoy a steady uptake, following the positive PEGASUS data. And the strong diabetes performance was driven by the well-executed product launches and the benefit of the global AstraZeneca footprint. Our emerging markets business showed notable strength in China and also in key markets outside China. And finally, Japan, you see at the bottom there, maintained growth in market share and product sales in what was a competitive environment. If you could just turn to the next slide we've got an addition. You can see from our announcement that we issued this morning, New Oncology has been added as our sixth Growth Platform. To ensure clarity, New Oncology is defined as worldwide sales of Lynparza, worldwide product sales of Tagrisso and also US product sales of Iressa. And of course naturally as we launch products such as durvalumab and tremelimumab, these naturally will be incorporated into this measure. So these medicines are instrumental in driving the next phase of growth in the years to come. Next slide thanks. So if we start on Respiratory, the franchise grew by 7% in 2015. This was driven largely by a combination of emerging markets and also new products in the US and EU. Symbicort itself declined by 3% in the full…

Thanks, Luke, and hello everyone. I'm going to spend the next few minutes taking you through our performance in 2015 as well as our guidance for 2016. Please turn to slide 17. Total revenue grew by 1% in the year, marginally ahead of our revised guidance. Our investment in R&D was supported not only by the growth in the top line but also by a one percentage point improvement in the gross margin and a 2% reduction in core SG&A costs. As you may remember, in 2015 we committed to a reduction in core SG&A cost by absolute value and also as a proportion of total revenue. Our improved R&D productivity and the increased focus on the main therapy areas meant we could also deliver over $1 billion of externalized revenue and $1.5 billion of other operating income. Further down the P&L, core EPS was $4.26, up 7% on the year which included a growth of 22% for the fourth quarter. The board remain committed to a progressive dividend policy and have declared a secondary sharing dividend of $1.90 per share, bringing the dividend per share to the full year to $2.80 for the full year, in line with previous years. As, we guide at constant exchange rates, and we anticipate a decline in total revenue in 2016 by low to mid-single digit percentage. We also anticipate a decline in core EPS by a low to mid-single digit percentage. The above guidance incorporates the dilutive effect arising from the Acerta Pharma and the ZS Pharma transaction announced late in 2015. Further, it is important to note that we may see greater fluctuations in the quarterly earnings performance this year, as a result of the anticipated loss of Crestor exclusivity for in May. I'll take you through more details on our…

All right. Great. Thank you, Marc, and hello, everyone. Please turn to Slide 24. 2015 was a good year for AstraZeneca. This slide highlights the key milestones including phase 3 readouts, regulatory submissions and regulatory approvals. The favorable outcomes are colored in green. The unfavorable ones are in gray. Clearly, the favorable events far outweigh the unfavorable ones. I will speak specifically to one of the gray boxes, which is SAXA DAPA and the complete response letter. In the first half of 2016, we now expect to make a new U.S. NDA regulatory submission for the fixed dose combination of saxagliptin and dapagliflozin. This decision is based on recent positive interactions with the FDA. In essence, we plan to submit additional clinical data for SAXA DAPA from a trial that is now completed. The key highlights were the six approvals including two new medicines, Tagrisso and Zurampic. We look forward to continuing this momentum in 2016, to deliver on the pipeline and keep you updated on our progress. Please turn to Slide 25. Next I would like to review some of the late stage pipeline highlights in the main therapy areas during the fourth quarter of 2015 and early into this year. Starting with RIA, the Symbicort LABA safety post marketing trial was positive. Zurampic received approval in the United States and a positive CHMP opinion in the EU. Brodalumab received regulatory submission acceptances into U.S. and EU. And exciting anifrolumab phase 2 results in lupus were presented at the ACR conference in November. In CVMD, Brilique received a positive CHMP opinion in the EU for the post-MI indication and ZS-9 received regulatory submission acceptance in the EU. Finally, for oncology, we received breakthrough therapies designation for Lynparza in particular forms of prostate cancer. Tagrisso was approved in the US…

Thank you, Sean. To please turn to Slide 20, I will quickly summarize the results today. Very quickly so we move to the Q&A. Revenue was up 1%. Growth Platforms importantly were up 11%. Our core EPS was up 7% which reflects that we delivered on our commitment to reduce SG&A costs, and we were able to maintain the momentum in the R&D investment. So our guidance is low to mid-single digit percent decline for both revenue and core EPS, and the dilution is of the acquisitions we made late last year is included in this guidance as covered by Marc a minute ago. Sorry about this. This microphone is falling. So I think really what this reflects is that we are on track with what we told you we would do. In fact, by and large we believe that we are ahead of what we think -- we thought we would do two, three years ago from a pipeline viewpoint. The only new development that we certainly didn't expect, I don't think anybody would have expected quite frankly is the negative development in the last 15, 16 months on the currency front, and this really represents a massive headwind for us like for many other companies that report in U.S. dollars compared to what the exchange rates were in 2013, for instance, we're losing more than $1.5 billion of profit in 2016. So the fact that we are able to still deliver the 4.20 [ph] at actual exchange rate in 2015 is a reflection of the strengths of our business. If we turn to Slide 30 before we end just wanted to reflect on our journey. We just finished the first phase of this journey which started early 2013, rebuilding our pipeline, and we believe now that we are…

Maybe I'll kick off with where you left off. The 4.20 floor has been sacrificed within reported guidance. Just wanted to understand what you felt that signaled, is it greater investment? Is it less certainty in one-offs? Is it just FX? Given the lack of earnings floor, potentially for the market now, when do you see profit earnings?

Yes, I mean -- I don't think we said that -- it's a good question, actually. But I don't think we said there's no earning floor. I think we've guided to the range of outcomes that we see are possible, but we're still very much committed to delivering on this core EPS goal we have, and I think really we should completely separate guidance from what our goals might be from a compensation structure viewpoint those are separate issues. The guidance actually reflects what we see for moving forward and includes the dilution from the acquisitions. As I said earlier, we still very much have committed to this EPS goal. The biggest issue for us is actually the currency. That's really the biggest thing we cannot predict. And so at constant rate we believe we are around the 4.20 and we certainly very much believe we're going to work hard to achieve it. The question for us is really currency. Hopefully that answered your question. If not you can come back to it and ask it differently.

I'll ask it slightly differently. James Gordan from JPMorgan. Just going below the 4.20 and below the 1.5 times cover, should we think of it as a one-off factor for this year because of the deal dilution or is it something that could be quite sustained, because it looks like 2017 is probably a tougher year than 2016. You've got a full year of no Crestor, a full year of no Nexium. I don't think the new launches are going to make that much difference by then. So is it something where you hope to just get a one off dispensation for this year, or could it be the new normal that we are below that level?

We haven't said -- we have not talked about dispensation. We've given you guidance at constant rate and certainly our goal is to bottom out this EPS. 2017 is clearly still a challenging year. Having said that, we believe that some of those launches will rapidly generate additional profit. We also believe that as we move forward and as the sort of launches, if you will, of products like Brilinta, diabetes and others matures, then the need for SG&A is going to be less as a percentage of sales. So we believe we can still manage the SG&A down. We think there is substantial reduction in SG&A for us to you achieve, the question is at what speed do we achieve this. And to achieve it too fast may actually threaten the top line goals. But we believe over two year, two and-a-half year period, we certainly can achieve substantial reduction in our SG&A. And that's really what we're going to manage to try and protect the EPS. And the way we protect EPS will be a mixture of as we've said before, a mixture of SG&A reduction and delivering on this externalization growth.

Assam Manyeed [ph] from Morgan Stanley, so another question on the guidance, I'm afraid. The credit rating, does that assume that you have reached the limit in terms of potential dilutive deals, and so you now if you want to do something it has to be accretive and no more dilutive? And just back to the SG&A comment, you reduced by $1 billion SG&A spending in 2015. Is that ballpark level of $1 billion achievable again in 2016?

Marc, maybe you want to cover the credit rating. SG&A as we've told you before, we're not going to give specific numbers. But you can see the SG&A as a percentage of sales, we do recognize is relatively high, relative to our peers. So the question is not so much can we reduce it, because the answer to that is yes, we can reduce it. The question is at what speed do we achieve that without impacting the top line in the near-term. That's really what -- we're on a gliding path in term of managing this SG&A, but we certainly will reduce it over the next two years at a substantial rate. Marc, do you want to cover the....

Let me position the credit rating and the dilutive impact of the acquisition at the end of last year. First of all as you know very well, the credit rating agencies tend to have an horizon of planning, which is more short-term than maybe equity investors. We need to understand the rationale behind these two acquisitions, and when I mention the two acquisitions I mean Acerta and ZS Pharma. In the short term, and we have indicated there would be moderate short-term dilution, but we have also emphasized the very strong impact on the operating leverage in the longer term. And that's obviously the rationale, the reason why we made this acquisition. So if you are looking at the credit rating, you would probably look more short term, but we look at the long-term perspective for equity investors. And then the more exact question on the rating, you know that the credit rating agencies downgraded us by one notch for both the two main credit rating agencies, but this does not mean, the reason why we look at accretive opportunities is not because of the rating, it's more because we also need to protect our cash flow and protect the other balance of the Company.

Also because we believe we finished, we have rebuilt our pipeline to a great extent. And really the focus has to be on execution progressing it delivering it, and also launching those new products. This is really where our efforts are. Thomas reminds me that I should ask questions on the phone as well. If you'll allow me I'll ask one question on the phone and then return to the room in a minute. Andrew Baum at Citi, do you want to ask your question?

Question for Marc on your tax rate outlook for next year, obviously the contribution from U.S. based products will be diminishing, a function of Crestor. And I assume that product impacted by the patent box probably increasing. So, should we expect there's any downside to that current 21% forward tax rate for 2016?

I don't know if it's at your end or our end but it's hard to hear. I don't know, Marc, did you get the question? We didn't get it.

Can you hear me now?

Yes, it's a bit better, yes.

Okay. Let me try again. The questions was that core tax rate for 2016 given you're losing the Crestor revenue in the U.S. with assuming a higher tax rate, and given the patents of products impacted by the patent box should be larger for next year. Should we expect there's further downward pressure on the 21% core tax rate that you posted for 2015?

So, I'm not absolutely certain I understood the totality of the question. But I think you're asking the guidance or the indication I provided on the range of tax rate, corporate tax rate for 2016. And I think you're asking whether the 16% to 20%, which we provided in 2015 is the same for 2016. So, I just repeat what I said in my talk, that we expect the rate, the corporate core tax rate, to be between 16% and 20% in the same manner that it is for 2015.

Although it is correct that our new products, many of them are in the patent box, and we'll derive a benefit from this from a tax viewpoint, but they are relatively small in the great scheme of things for the time being. But certainly over the next few years maybe a little bit this year but mostly the years after we'll derive a benefit, but it's probably a little bit ambitious to expect the tax rate to drop in 2016 because of the patent box. I think 16% to 20% is probably the range that you should expect.

I think I mentioned the caveat in my presentation that obviously the core tax rate depends on the geographic mix of where the profit is realized as we need to pay taxation in all countries where we generate profit, so within that range. Alexandra do you want to.

I have a completely different question. On slide 16, you did show some relatively steady uptake of Tagrisso. And I look at the IMS data it looks a bit different. It looks like a very, very fast uptake in the first three weeks and a bit more steady then. Independent whether this data shows the right thing or not, the question I have is what do you find in the marketplace in terms of rebi -- the challenge of having to biopsy mutations? How widely is that done? Is there a long way to go? You mentioned a BRCA status how much there was to go, and can you tell us roughly where you think we are in all the three key territories?

Yes, sure. First there's a high degree of understanding, understanding of the product, I think particularly in the US because of the dynamics there, around 10% is the testing rate for T790M mutations in the US right now. You're right, the key challenge is that second biopsy. But there is actually, the feedback that we're getting, and again it's early days that there's a high enthusiasm and interest in doing that and that physicians can explain that to patients and make the case to do that, because they themselves are convinced. Of course if we look in the midterm, the promise of the CTDNA test is attractive, because of course then you're really focused on resolving the false negatives. And you just might lower the barrier to initiating that treatment. And, ultimately, of course if we can have data on first line then it's a simpler discussion. Net-net very positive support for it. I think the reality is in the US we have a much cleaner area to operate than we were expecting, and we really have the market to ourselves for some time.

[Indiscernible].

Yes. The thing to keep in mind of course is that BRCA testing in the US is a relatively high rate because of course there's a broader consequence of the test which in terms of family members. So there's a higher expectation and demand for the test. Historically with the T790M test there was no clinical strategy that can be derived from that. Now we've supplied that. We have three labs in the US, which are centers of excellence so the infrastructure is in place. Historically we've had experience with EGFR testing. We'll now do you that through Europe and also emerging markets. We're very actively involved in that infrastructure in Hong Kong and Macau. Sorry?

[Indiscernible].

Japan as well. We've got a team in place right now which is a very, very good team, tiding lung physicians and of course remember we had a team in place which is really why we brought back Aresta in the US which essentially have a you running start for the products.

Trying to follow some other. Hopefully I get it right.

Thank you. It's Matthew Weston from Credit Suisse. Hopefully the last guidance question. You were very kind in setting out expectations for externalization and thank you for that. But in 2015 you generated about $1.3 billion from product divestitures included in other operating income. If you could set out what's your expectation for product divestitures in 2016 that's baked into guidance. And within that, have you identified the products for sale? Have you started discussions, or it's just an aspirational goal that that will help you modulate the P&L this year?

I would like to bring you back to 2015. I believe that at the same in a different place but the same time last year we were having the same level of discussion on how are you going to deliver your guidance for 2015. I think I answered then that it would be a conjunction of two factors, the reduction of SG&A expenses, and I think you now see that we have delivered on this reduction, both in absolute terms but also in percentage. And I also said at that time that we would generate revenue through, externalized revenues, and the question was raised then what percentage of both, and my answer then that you would be relatively two similar factors, one would be at 10%, 90%. I think you can use the elements of 2015 to project to some extent for 2016. One indication I can provide is that in 2016 there will be more of it. If you take the total revenue, the externalization revenue, the total of external revenues and other income. So we're going to be a more pronounced effect for this impact.

So just to clarify, if I take the billion of externalization income in 2015, add the $1.3 billion of onetime other operating income, it will be greater than that $2.3 billion, however it's made up of the two buckets.

Absolutely.

Thank you very much.

Maybe the one thing I would add is that we are pursuing two strategies that are really all driven by focusing ourselves. And the big one is of course externalizing, partnering some of our products, and I think the other one of course is divesting assets that are small. And if you actually look at it, we have generated a lot of income out of products that we're selling $40 million to $50 million and we're sort of not doing much in our hands because we're focusing elsewhere. But the one important point is people see this externalization revenue sometimes as one-offs. Essentially what we're doing here is sort of priming the pump if I may. Maybe it's not the right expression. Actually we're starting a process of introducing a new business model, but at some point this will turn into recurring revenues. If the base inhibitor delivers, if it gets approved and it's launched, the externalization revenue line would be enormous. And so there is quite a number of products that will actually be like this, that will continue to deliver revenue on an ongoing basis. Essentially what we do is we say those are products we don't want or cannot develop ourselves because we don't have the capabilities, we don't have the financial resources, we are not focused. But we want to retain some economic interest. We partner with someone who's going to do a good job operationally for us and then leave us with some of the economic value. Should not really be seeing that as one-offs. They're of course one-offs but there's also long-term recurring revenues that will develop.

Thank you, its Mark Purcell, Barclays. Can we just follow-up on that, and we'll ask something totally different. If you take externalization revenue, around $1.1 billion for 2016, can you help us understand the mix between recurring which you were saying is going up, the milestones, we know of one, obviously 170 on the Brodalumab. And then the new stuff that we don't know anything about. And then slightly different on durvalumab and the updates on the pipeline, can you help us understand why you chose on the triple to do durva plus chemo plus or minus treme why you didn't do durva plus treme plus or minus chemo? And then secondly you changed the primary end point on mystic as well. Was that done because you're concerned about patient crossover in that study?

Sean, do you want to cover the last two questions and Marc the first one?

So the first question, I'm going to the second one first and then I want some clarification on the first one, okay. So the second one was change of end points on mystic. The change just so everyone he knows it's making a coprimary. Progression free survival, the previous sole primary, and combining it with overall survival. This really came from two things. One is as data emerges with immuno-oncology, I think we are wondering about progression free survival as the best indication of the benefit that patients derive from this particular mechanism of action, because we have seen -- it's a mixed bag right now, and its early days. But we've seen some places where they seem not -- the progression free survival seems not to really predict very well an overall survival benefit that comes through. So that led us to think we should move the overall survival up in the hierarchy. The first analysis will be progression free survival. The second aspect of it really is a pragmatic one. And it has to do with how quickly the trial is enrolling. So one concern when you do this is you have to add some patients to the trial, because you want to power both your primaries, and we did that in the amendment. The thing is that what we had forecast, how long it would delay us turns out to be a fraction because it's enrolling so quickly. So we were able to do this change without really impacting the time line. So those were two of the major factors that went into it. Now, the first question, can you clarify for me for a second?

Sure. Sorry. Obviously there's a big focus on durva plus treme, but the triple trial is durva plus chemo, plus or minus treme.

That's right.

I would have thought it would have been durva plus treme plus or minus chemo, sort of put the variance on the chemo to change the.

Let me talk about two things about our fundamental strategy. You could do it either way, you're right, it depends on what your fundamental strategy is. Our emphasis is combination, combination immunotherapy, I should say rather. There's a little bit of a pragmatic aspect of this. Some pragmatic and practice patterns doctors treating lung cancer, some actually pragmatic in the world, availability of therapies. We've heard very loud and clear from some fraction of physicians that, hey, I'm going to give my patients doublet chemotherapy. It's shown a survival benefit. It's well established. And if I can combine in, great, but I don't feel I can deprive them of the chemotherapy. What we're doing is we're trying to provide, one, for the combination strategy and two, a data set for that not insignificant group of physicians to be able to understand what's going to happen.

Essentially that study is focused on the combo with chemo with the other dimension to test whether adding treme on top without something whether we really want to establish durovet [ph] chemo, the benefits of that combination. Should we move -- sorry, Marc.

I think I'm going to answer the previous question on do we have released or do we have a concrete plan for revenues for the 600 [ph] revenues or divestment the answer is yes. I'm not going to give you the list, so please do not ask. But I'm just going to explain why we have a list so we know which product, which activities and also which parties could be interested. What we don't know yet, although we have already ongoing discussion, what we don't know yet is what structure that it will be like. Obviously it depends on the counter party, and this is why it is not easy for us to tell you what part will be recurring, what part will be sort of one off, what part will be in external revenues and what part could be in other incomes. It obviously depends on the structure of the transaction. But I think if you look at the totality of external revenue and other income and you see an expansion versus the level of 2015 I think you're in the right place.

Should we move to the online questions again? Tim Anderson at Bernstein. Tim, go ahead. Go ahead.

Thank you. Your first line lung development program with durvalumab, it's pretty apparent you're putting almost all of your eggs in the basket of combination therapy with tremelimumab. But I'm wondering what happens if the combo data ends up looking lackluster. Your front line monotherapy trials are in all comer lung patients, but your competitors seem to feel that in front line lung it's risky to look at unselected patients so they're only looking at mono therapy front line in biomarker positive patients. So can you clarify for us what happens to the durvalumab franchise in front line lung if the treme combo ends up not working?

Sean, do you want to --

You added franchise in the end there, so I might absolutely get the end. But with regard to the outcome, but let's go with the outcome of mystic which is really the lead front line trial. As you're alluding to, Tim, that has durva treme, it has durva as a single agent and then it's versus standard of care, so three arms, one to one to one randomization and now PFS and OS as end points. And we are taking all comers. Obviously the data will be analyzed by biomarker positive versus biomarker negative, and our hope is, we've presented some data that we feel supports this, that durva treme will be you truly differentiating in the PD-L1 negative patient population, which I will add is the majority of patients, two-thirds to three-quarters patients depending upon what data set you look at. With regard to the PD-L1 positive population, it is possible that durva and treme and durva and are both active and active in a way that can't be differentiated. In which case it is possible for us to file based on that data a single agent durva for that PD-L1 positive subset. So I guess what I'm saying in part is it depends upon the data, but as you point out there are multiple opportunities or permutations that's could arise depending on what it shows.

Does this cover your question or --?

Yes. I guess the question was whether there's enough powering in that monotherapy arm of mystic just in PD-L1 positive patients to use that as a registration subset, but you're saying that it is adequately powered to kind of carve of that data out and if I could just clarify. So your view also, and I think you've said this before on the percent of patients in front line that are PD-L1 positive versus negative, that's a figure that differs from what Bristol-Meyers has said. I guess I've always tried to figure out why that difference exists.

So I'll tell you one challenge we have. Hopefully this will be resolved. And I think some of what you're looking at is assay and assay cutoff effects, and now that the Bristol-Meyers and the Merck assays are out there and can be used for comparison, that presents an opportunity to really look at the same samples with the different assays and help to resolve this a little bit. But you're right, there is a range of variability. PD-L1 negative is the majority, but you're right, the exact numbers do vary, and I think it's probably dependent upon the assay and the assay cutoff mostly.

Thank you.

Thanks, Sean. Seamus Fernandez, maybe we could ask the other question online here. Seamus, go ahead. Then we'll return to the room in a second.

Great. Thanks. So can you hear me okay?

Yes.

Great. So just wanted to ask specifically, Pascal, as we take into context the opportunity that you're looking at, potential accretive acquisitions, would you be willing to kind of put some brackets around the type of acquisition that you would consider? There's an awful lot out there. And you how long do you think it will take for some of those acquisitions that you might be considering for the owners of those businesses you think to kind of come together with you, given the rapid pace of decline that we've seen? I assume that you would be more looking at biotech type entities that are strategic. But maybe if you could just give us some brackets around the types of acquisitions that you would consider and maybe relative size as well would be really helpful. Thanks.

Okay. So thanks for the question. The first thing is that as we said, they have to be accretive because, again, we believe our pipeline is full now. So they have to be accretive. They have to be the right price. Importantly, they have to be strategically aligned with what we're trying to do. So they have to be in autoimmune, respiratory, in cancer and cardiovascular diabetes, so that we keep building our presence in those key therapy areas. And finally, the size really depends very much on how big is the cash flow that is added through the direct acquisition and through synergies. It depends on our ability to -- how much we can raise money as a result of all of this. As I said before many times, we're agnostic as far as size, provided it's strategically aligned, the price is right. We can add value, and we think we can execute. Executing on these acquisitions is critical so that we don't distract the organization to an extent that what we gain through the acquisition we lose through distraction of the pipeline. So that's really as specific as I can be. In terms of getting together, again depends on the size. I mean, some of these acquisitions we integrate very quickly. We can integrate quickly, and the bigger they are the more complicated it is. But again size is only one consideration, the geographical complexity, et cetera, et cetera. Every case is different, really.

May I ask one quick follow-up on a clarifying piece of information on one of your pipeline products? Roxadustat in China, will we -- do you anticipate that we'll actually see data in a public forum based on the data in China or in 2016, or would you anticipate that sometime in 2017 because that trial will wrap up very late in the year? Thanks a lot.

So that's something that ironically, probably is something that's more sensitive for everyone in that group. But again, good start and also the flip side of that of course is we lost some access with Bydureon, and we took a defensive posture, and I think we'll come out of that quite well.

I think it's an important point that Luke is making. If you look at the prescription share, the increase is very impressive. It's kind of nice but you should be careful to not extrapolate this too rapidly because you never know how things settle down. Certainly it's a very encouraging start to the year, there's no doubt about this. And the bigger issue is really the class. If you look at the total class, it's relatively flat. Essentially out of the ketoacidosis issues it's starting to pick up. We have good hope that the class will pick up. In fact if you look at Japan the class was negatively impacted also initially for different reasons. It was really the hydration issues in Japan. But the class is starting to pick up quite nicely. This is a good class, and many new agents face issues at the beginning, just if you remember Crestor, some of you may not remember Crestor, but Crestor faced a very challenging start, again for septic questions. The Company at the time was able to resolve those issues, and the product did extremely well. So this is a good class. It's going to do well. There's no question. You just need to work through the issues in Japan and the U.S. but for us in the short term that's really the biggest question is how quickly will the class pick up. Then we'll come back to Marc in a second.

I'm afraid I am going to bore Matthew with two calls from me on the guidance. Firstly, you've included within the guidance the dilutive effect of Acerta and ZS Pharma. I was wondering if you could quantify that in aggregate. And secondly, to what extent is the Q4 SG&A performance a guide to 2016? Thank you.

We are not going to quantify the dilution. When we did the acquisition, the separate acquisition, we indicated the type of dilution. We think we said short-term and minimal for ZS and we said moderate for Acerta. And we also said short term. I think with this sort of compilation of adverbs and adjectives I think you can probably derive the impact of this dilution. I'm sorry but I'm not going to give you the number, the exact number.

The only think I would add is the range of guidance at constant rate gives you a sense of where we think we could land and also of how do we manage the dilution overall. But we can't give you a specific number for sure.

We mentioned I think I mentioned in my talk, just to complement my answer, that we will be taking 100% of the R&D of Acerta, but as we only own 55% of the equity of this company, if there is a loss because a product is not yet sold, 45% would flow back, 45% of the loss would flow back through the minorities.

Again that minus 11% in Q4 is it an indication for 2016.

Well, I said the trend would -- the trend of 2015 on the whole year will increase but probably not reach 11%. If I tried to guide you, we've given you this guidance at constant rate for 2016. We actually -- I'm personally very committed to delivering what we said we would do before. And essentially what we would do before in a roundabout the ways is we're going to bottom out in ‘16, ‘17 and grow from there. We're committed to this. I'm committed to this. The only thing I can't absolutely forecast is the currency. That is the one thing that is really out of our hands. And that currency impact we will do our very best to compensate some or all of it. In fact, we're hoping the currency impact will go the other way and help us. But if it doesn't help us, we'll do our best to try and compensate. We can't be sure we will be able to totally compensate for it but we're very committed to delivering, I'm committed to delivering what we said we would do before which is sort of bottom out over the next two years. This year or next year, it's no longer the next two years and then grow from there. Matt?

Thank you. Luke, can I ask a roxadustat question? Clearly we get the Chinese data this year and it could prove to be a very exciting start for the whole class. But can you just remind us on timing of the current approval time in China in terms of how long it's actually taking on average for a drug to get filed to approval. Because we think of it as 12 months in Europe and the U.S., but I recall it as potentially significantly longer than that. When should we think of the commercial opportunity for roxa?

Sure, just in terms of tracking approval time frames in China, I've seen so many reports on this and have tracked it myself. It's almost impossible to do the analysis because the key thing for roxadustat is it's actually treated as a local product. So it's produced, the material is in China, it's actually a different formulation than the global supply. So we said the correct terminology for the Chinese submission for roxa is rolling submission.

It's rolling submission. To that end, what we said is the submission would initiate this year. We did not say there would be clinical data available this year.

The thing to take home, it's very difficult to plan on Chinese regulatory time frames. However, clearly it's a product which because we've contracted to file there, there's a dedicated program there in China. And I think you could expect that obviously the profile of the product has to hold up but it should [Technical Difficulty] in China is enormous, both in terms of the straight out dialysis population. Also peritoneal dialysis is quite common in China as well. So again, there's some attraction there. And then if you look at the pre dialysis population that's directly correlated with the disease burden overall with such things like diabetes and that population in China is unfortunately quite large. So it remains an attractive product.

The process is not as codified as say a rolling submission in the United States, but it's still -- the product is still part of a new process called the green pass. And clearly the authorities have put it in this process because it is a local product as Luke was explaining. And it's really addressing an enormous need in China. Diabetes is exploding. Kidney disease is rapidly growing. They don't have enough dialysis centers. They need new medicines to help had those patients. So clearly, they have a lot of reasons to fast track this project, but it's not very codified. And so it's hard to predict how long it will take.

Just a very quick follow up to Sean. When will we see clinical data if it's not this year?

I don't recall what we committed. Yes, I think it was half one next year was what we committed to. Let me look that up for you.

While you do that, maybe we could move to another question and return to that. Little bit far from this side.

Carrie [indiscernible] with Shmecksa, I have two questions on respiratory please, so firstly on Symbicort. Clearly, the pressure is stepping up in Europe. So your thoughts on how that will evolve from here. But also your expectation in the U.S. market, potentially generic, that we can debate but how you think about that, and what you put into your guidance, your expectations there for Symbicort. And then secondly on PT003 is there anything that you can learn from GSK's experience with ANORO and clearly now they struggled to take share from spiriva, and they appear to be reverting back to pushing that single agent LAMA in combination with rio to offer the open triple, which you won't have the opportunity to do until you have your closed triple. What can you do with PT003 in the meantime? Thank you.

I'll break that into chunks. With Europe, we've got three analogs there. So we've kept 90% of the volume. So it depends on the market. Some of these markets you have an automatic mechanical drop in the price. You're out of their control. Then you're arguing over volume. We've maintained that volume effectively. So I don't -- we're certainly we don't guide by product levels, but there's been some figures floating around out there. I think it's fair to say we have -- we're quite confident about that market. There will be pressure on Symbicort in Europe. But again, we're quite comfortable with where it's going as a trajectory and where it's holding and our ability to hold that. Remember we've been able to grow share with Symbicort globally. We've actually grown share in the U.S., too. If we look at the U.S., there's price pressure. But again, volume does compensate for that. We've picked up MBRx. We've picked up TRx. So in a marketing battle, I think we can keep share and hold it. So there will be price pressure but not catastrophic price pressure. In terms of generics in the US what's being disclosed publicly is exactly in line with what we had in our own internal forecast and what we built the long range plan on. And that's filing of course. We have to see if they get approved. We know this is quite a complex history with inhaled molecules in the US. In terms of changes in the dynamics of the market, I think in some levels there's a timeframe there for the payers so they've got this on the horizon so it actually changes the dynamics in terms of the trade off the companies are prepared to make. Again, we look at each negotiation in terms of that balance between supporting value and not destroying value and then also market share and try and make an intelligent decision out of each one of those events. In terms of PT003, maybe I just focus on I've said this before with saxa dapa. It was a little bit longer than we were hoping for but there is an advantage in coming second sometimes. You can learn. I think there's broader dynamics that now that BI has entered the marketplace with their LABA/LAMA that's starting to change things before with bialto that he may shake things up a bit more and dislodge some of these patients who are on a LABA and who are quite sticky. But there's lots of things we can learn in terms of preparing the market, doing some work there, and we've shown in Europe with our other LAMA/LABA that we can take share and compete. So with 003 I think we're quite confident. Ultimately, it depends on the label you get. But we're optimistic, and we think there will be a high demand for this product in that formulation.

It's one of those classes that takes time to develop, but it's a very dynamic market, and the problem is that the share of dynamic patients is actually limited in the total prescription volume. But if you look at the dynamic market itself there's a certain number of countries, 40% to 50% of your new scripts, new initiations are for LABA/LAMA away from spiriva. It's actually starting to pick up in quite a number of markets. The problems is that it really takes time for those new initiations to turn into a large volume of total prescriptions. Should we go back to.

Let me go back to the roxadustat. It's 2017, but again, the submission is rolling, it's ongoing. As Luke said, we're not 100% sure what the review time is. It is not codified. 2017 there will be data from trials outside of China as well.

Marietta Miemietz, Prime Avenue. Just wanted to make sure I understand your SG&A comments correctly. So presumably most of the cost that you're looking to take out over the next couple of years, that's actually primary care infrastructure which you can basically take out as soon as the loss of exclusivity occurs and not a minute earlier. And then you just have a three to six month lag effect before it actually comes out of your core P&L, because it takes some time to get down to brass tacks with the employee representative. Is that the right way of looking at it in terms of modeling, or do you actually think you can make any significant improvements to marketing effectiveness in emerging markets or specialty care, and if so, how would that come about? And then just a very quick question on the Tagrisso Japanese label you're expecting given that the Japanese are really known for granting quite broad labels in oncology and not so much dissecting between different segments.

Sean, maybe you can cover the second question. Let me cover the first one very quickly. The SG&A savings they will they're coming, and they will come from a wide range of sources. We work a lot on admin cost, IT cost. We reduced our goal was to reduce by 30% from 2014 to 2016. We've gone quite a bit we've gone quite a long way. We still have more to go this year. We're working on site costs. We're working on a variety of G&A costs. When it comes to pure commercial costs, you're right, that there is quite a bit of cost reductions that then come out of reducing ifelt and primary care over time. But we also have a program to improve productivity and efficiency in our commercial teams in all the emerging markets. We have an entire program that is focused on that. We've increased a lot the size of our teams in China in a number of countries through the acquisition of the diabetes business but also you through expansion of our teams. We're now focusing really hard on improving the productivity of those teams. So there's a whole range of sources of cost savings and productivity improvement. We come back, can we have one or two more questions? Okay.

You want me to answer the Tagrisso

Sorry, the Tagrisso Japanese label.

It's a little hard to speculate on the label but regulators do what they want based on what they see in the data. The observation I will make is the EU label is different than the United States label. In the United States we have the T790M. But then you have to have failed first generation EGFR molecule. In the EU the T790M is in there, but this first generation having previously been treated is not in there. Japan does tend not to make little cutouts in its labels. So we'll see what happens. Can't really be sure.

Sachin, maybe and that will be the last question, Thomas, or can we? Okay.

Thanks for taking my follow-ons. Just one on cash flow. Given the vagaries of the P&L, wonder if you could comment on cash flow. Operating free cash flow was roughly $2.5 billion and didn't have the dividend this year. And that included $1 billion of externalizations. Just any color on free cash flow will cover dividend in 2016 and how we should think on a two, three year view, given the progressive dividend policy. One for Luke on Brilinta. You mentioned in your intro comments that an inflection might be seen post guideline updates, so any color on when we could expect that?

On the Brilinta side, we have actually seen already an inflection in the DDDs in hospitals, and that reflects a pickup in initiation. But, Luke, do you want to cover this and, Marc, you cover the second one.

Sure. And right now if we look at volumes on the 60mg, it's still relatively early days, but the DDD really did jump. And if we look at the data in terms of who's being put on 60mg, we're pleased to say that 70% of those patients are actually beyond 12 months or around 12 months for Brilinta. In terms of guidelines, it's difficult to speculate. We hear certain things. We have asked questions. You've got the ACC of course and the AHA coming up, so around key Congresses like that where alignment seems to be there. I would direct you back to the label in the US with the statements around clopidogrel. And just the broad nature of the label which again gave us some confidence. But we'll see what we get. We ask lots of questions about it.

Turning to the cash flow, basically you can -- the level of net cash flow will be similar in 2016 as they are in 2015. We need to understand the contributory nature of the externalization revenues but also on the other income. It does contribute to -- obviously to our cash flows, and it helps us sustain a relatively high level of R&D. One needs to understand that. External revenues is not only to sustain the cash flow, but the cash flow can be reinvested in R&D. I think the whole cycle needs to be understood.

The cash flow was the same in 2016, more or less the same as in 2015. But I think Marc is raising another point which is an important one. Externalization revenue is helping us to create long-term value by partnering with someone who will turn a product that we probably would not do so well ourselves with because it's not part of our core strength. So long term value but also short-term value because some of that money we are investing immediately in building a strong pipeline. We could save money and still deliver the same EPS with less externalization, less cost, would we create a better business long-term? Probably not.

Thank you. Your remark is true for the year 2016. It will be very similar to 2015. But you know we also need to see that -- we have provided in early 2014 a signpost for 2017 where we said basically the revenues would be more or less broadly in line with 2013 at constant exchange rate. I think you can just extrapolate from this in many, many positions the P&L. What we are doing is increasing our pressure, our cost discipline and our pressure on SG&A. But we have conversely also increased the R&D spend.

With that one I could call this meeting to a close and then thank you for joining us. But in parting, just let me wrap up leaving you with a few messages. First of all, we're very much committed to the dividend. Nobody should ever doubt our commitment to the dividend. Two is we're very committed -- I'm committed as I said to delivering what we told you we would deliver and bottom out in 2016, 2017, grow after this. Our guidance is in the range of this. And hopefully the question is not whether we deliver plus, minus 3% or 5% on this core EPS in 2016, 2017 but what kind of long-term value creating for the pipeline. We're very committed to this bottoming out. Again the thing we don't totally control is currencies, currency movement, and we'll do our best to compensate some of those movements if they're negative. If they are positive, we'll just welcome them of course. But we can't guarantee that we will be able manage those. But certainly very committed again to managing the 2016, 2017 period of time. Again, I think we're making great progress through the pipeline. We're launching our new products. Tagrisso is doing very well. Really, hopefully, we can focus ourselves on the post 2017 period of time. In the meantime, through this externalization and cost savings we really are defending our EPS, and hopefully at some point everybody will agree that externalization did work and we created long-term value.

With this thank you so much for joining us today.