Welcome to the first quarter 2016 Novo Nordisk earnings conference call. For your information, today's conference is being recorded. At this time I would like to turn the conference over to Mr. Lars Rebien Sorensen, CEO. Please go ahead, sir.

Thank you very much and welcome to this Novo Nordisk conference call regarding our performance in the first quarter of 2016 and outlook for the full year. I'm Lars Rebien Sorensen, the CEO of Novo Nordisk. With me I have our Chief Financial Officer, Jesper Brandgaard and Mads Krogsgaard Thomsen, our Chief Science Officer. Also present and available in the Q&A sessions are Executive Vice President China Pacific and Marketing, Jakob Riis, Lars Fruergaard, Executive Vice President Corporate Development and Vice Chair of our Operations Committee. Present are also our Investor Relations officers. Today's earnings release and the slides for this call are available on our website, Novonordisk.com. The conference call is scheduled to last one hour and as usual we'll start with the presentation as outlined on slide number 2. The Q&A session will begin in about 25 minutes. Please note, as mentioned, that this conference call is being webcast live and a replay will be made available on Novo Nordisk's website. Turn to slide number 3. As always, I need to advise you that this call will contain forward-looking statements. Such forward-looking statements are subject to risks and uncertainties that could cause the actual result to differ materially from expectations. For further information on the risk factors please see the earnings release and the slides prepared for this presentation. Turn to slide number 4. Sales growth in the first quarter of 2016 was 9% measured in local currencies and 8% in Danish kroner compared to first quarter of 2015. The growth was primarily driven by United States and international operations. Sales growth was realized within both diabetes care and biopharmaceuticals with the largest contribution coming from Victoza and Norditropin, the latter being driven by a non-recurring Medicaid rebate adjustment in United States. The rollout of our new-generation insulins…

Thank you, Lars. Please turn to slide 10. In March we announced positive headline results from LEADER, the cardiovascular outcome trial for Victoza. The trial was initiated in September 2010 as requested by the FDA and DMA with the aim of providing long term data regarding the safety and, in particular, cardiovascular safety of liraglutide relative to the current standard of care in the global population of people with type 2 diabetes. The primary endpoint of the trial was the first occurrence of a composite endpoint encompassing cardiovascular death, non-fatal myocardial infarction, also known as MI and non-fatal stroke, normally referred to as strict or three-point MACE. The LEADER population consisted of CV risk-enriched type 2 diabetes patients with an average age of 64 years and 13 years of diabetes duration. HbA1c was 8.7 at randomization and the BMI was 32.5. LEADER achieved its primary endpoint by first demonstrating CV non-inferiority and subsequently superiority for liraglutide on the primary MACE endpoint. This is the first GLP-1 analog and the second single anti-diabetic agent ever to demonstrate a cardio protective effect in type 2 diabetes. The MACE reduction in the liraglutide arm was driven by all three components of the endpoint, mortality, MI and stroke. From a general safety and efficacy perspective LEADER data were consistent with the already well-established profile of liraglutide as known from multiple clinical trials. The comprehensive dataset will be presented on Monday, June 13 at a symposium during the American Diabetes Association's 76th scientific conference in New Orleans. Furthermore, Novo Nordisk plans to submit all data from the LEADER trial as soon as possible in the form of a safety and efficacy supplement to the regulatory files for liraglutide. Please turn to slide 11. In the first quarter of 2016 we also announced the headline results…

Thank you, Mads. Please turn to slide 15. During the first quarter sales increased by 9% in local currencies and by 8% measured in Danish kroner to DKK27.2 billion. The reported gross margin decreased from 84.6% in 2015 to 84.4% in the first quarter of 2016, reflecting a negative currency impact of 0.2 percentage points. The underlying gross margin remains unchanged, driven by a positive impact from product mix and prices from higher Victoza sales and non-recurring Medicaid adjustments countered by ramp-up cost for new manufacturing capacity. Sales and distribution costs increased by 11% in local currencies and by 10% in Danish kroner to DKK6.7 billion. The increase in cost is driven by the U.S. and includes launch costs related to Tresiba, the continued rollout of Saxenda and NovoEight, as well as direct to consumer advertising. Sales costs have also been driven by sales force investments in selected countries in international operations. Research and development costs increase by 2% in both local currencies and Danish kroner to DKK3.3 billion, driven by higher research costs for diabetes and obesity projects. This is partly offset by lower development costs due to the wind-down of clinical trial costs for the cardiovascular outcomes trial, DEVOTE, the phase 3a program, SUSTAIN, as well as the SWITCH and LEADER phase 3b development programs. In the first quarter of 2016 the oral semaglutide phase 3a program, PIONEER, was initiated. This is partly offsetting the decline in development costs. Administration costs increased by 9% in local currencies and by 6% measured in Danish kroner to DKK908 million. The higher administrative costs are driven by higher costs across the regions mainly related to higher employee-related cost in international operations to support the growing organization as well as litigation costs in the U.S. Other operating income was DKK284 million in…

Thank you very much, Jesper and Mads Krogsgaard. In summary, we're very pleased with the performance of the first quarter 2016. Our key products continue to perform well. We look forward to the coming months in 2016 in which we will continue to progress on the launch of Tresiba in United States. We also look forward to be able to share the successful results of the completed LEADER and SWITCH trials in connection with the ADA and eventually also the SUSTAIN 6 data later in the year. We're now ready for taking the Q&As, where I kindly ask all participants to restrain yourselves and I would like to emphasize this - to two questions only to allow as many people to raise their questions as possible. We're ready to take the first question, please.

[Operator Instructions]. We will now take our first question from Michael Leuchten from Barclays. Please go ahead.

It's Michael Leuchten from Barclays. Two questions on price and mix, please. Firstly, on Victoza, if I use the IMS volume trends and the reported U.S. constant exchange rate growth rate I can back out the implied price and mix effect and since Q4 we're not seeing a positive mix effect here, also not in Q1. So is there increasing competition from the likes of Trulicity? Have we seen a plateau in terms of patients titrating to the higher dose? Any commentary around that would be helpful. And then similarly, for NovoLog, we've seen commentary out of your competitors in the U.S. in terms of the short-acting space. However, we're not seeing that in your portfolio in Q1 as far as I'm concerned. So just wondering whether that is the United Health contract you walked away from or whether there's anything else going on. Thank you.

In regards to the situation with Tresiba it is correct that Trulicity is doing well and this means that we're losing share in Victoza. But we're growing our volume with a double-digit growth rate so we're satisfied with this expansion and especially in as much as we're seeing an overall growth, very substantial growth in volume. We mentioned this to be more than 25%. This bodes extremely well for our entry into the once-weekly segment later on with the hopeful approval of semaglutide. In terms of NovoLog it is correct that our business is impacted by two factors in the U.S. It is a contract loss, a significant contract loss which reduces volume. I have to admit that the value was modest when we take the rebates into consideration and, therefore, we decided not to participate in the contract at that level. And then it's also significantly impacted by a negative influence from the rebate adjustments that we have discussed repeatedly where there was a positive growth hormone effect and a negative modern insulin effect. Thank you.

Maybe, Lars, I should just clarify that the impact on growth hormone was to the tune of DKK600 million and the negative impact primarily impacting the insulins was minus DKK200 million, so a net effect of DKK400 million in round numbers.

Our next question comes from Sachin Jain from Bank of America. Please go ahead.

Two please, one for Mads. On the Xultophy AdComm in mid-May, do you have any visibility on what is expected to be questions within that? And do you expect the same or different topics versus LixiLan? Second question for Jesper. On the headlines earlier and your media interview you'd cited biosimilars entering pricing discussions for 2017. Are you trying to signal any specific pressure there and do you have any early indications of pricing on the diabetes portfolio into 2017? Thank you.

Mads Krogsgaard with expectations or speculations on FDA's questions for the AdComm.

Yes. And at this point, Sachin, it will have to be speculations. We have not received the pre-meeting package or other material from the agency in this regard. Obviously, both IDegLira and LixiLan represent fixed-ratio combinations of two biological agents that together represent a new class of agents that don't exist. So it's rather natural that the agency when they see a whole new class coming of age will want to host an AdComm. And in that regard it's also relevant when you have a new class of agents to discuss what is the positioning and the labeling considerations for such an agent and, in particular, this combination agent or drug. So I don't have any specifics. I could imagine, though, that the two days will also be a subject of debate when it comes to more product-specific phenomena. And in regards to IDegLira I think you know the picture. I don't think we have anything that surprises us. It's really just a fixed combination of a well-established liraglutide molecule and a, being well established as we speak, insulin degludec molecule. So it's probably about the class being so new.

Thank you, Mads. Jesper, comments on your expectations for biosimilar impact on pricing and pricing in general in 2017?

The comment I made was really looking at the U.S. market going into next year and I think I comment that we have a relative stable situation within the rapid-acting and the mix in terms of market shares and bidding and I don't anticipate that that would significantly heat up when we go into 2017. Whereas my comment on the media call was basically just reminding everybody that it was expected there would be a launch of about a biosimilar version of glargine towards the end of 2016 and the implication that that would likely have for prices of basal insulin in 2017. Those discussion is expected to take place over the coming four to five months and, hence, I would assume that we have more clarity on that matter in - sorry, in the August conference call. I would anticipate that we would see some pressure on the basal insulin prices for 2017.

And then at point in time there will be five product entries into this category, some being generic versions of Lantus. Lantus [indiscernible], Levemir and Tresiba, so obviously this is something that the bio side, the PPMs, will try to take advantage and we will be having a much clearer picture of this as we get to half year and third quarter of this year.

Next question comes from Peter Verdult from Citi. Please go ahead, sir.

Two questions for Mads, SUSTAIN 6 and just touching on some of the early-stage research efforts at Novo. Mads, I know enough to be dangerous on statistics, but not much more, so I just want to make sure I've got my facts right. If you're showing a CV benefit on SUSTAIN 6 with the accrual of your 250 events, am I right in thinking that that could make the hazard ratio in SUSTAIN 6 as low as 0.7? I realize you're not going to give an answer to that in terms of what it is, but I just want to make sure, theoretically, could it be as low as that? And then just touching on some of the discovery and early-stage R&D at Novo, can you just remind us what level of investment you're making in terms of stem cell therapies. There was some data published by Harvard in cell animal data, albeit showing that you could transfer successfully insulin-producing beta cells back into rodents and induce normal glycemic control. Just wondered if you involved in that sort of area. And then just very quickly, on the Alzheimer's data with liraglutide, is that something you're going to pursue going forward? Thank you.

That was three questions, but we will take it anyway. Mads Krogsgaard, speculation on the specifics in SUSTAIN 6?

Yes. Well, since we have revealed the amount of strict MACEs to be around 250, of course, if I either gave you a P value or a hazard ratio then you could calculate the other one also. So what we have said is it is clearly statistically significant and that basically means that we can claim we're statistically significantly improving the cardiovascular outcomes on the strict MACE endpoint versus standard of care. And in principle the hazard ratio could be all the way down to zero if we had no events in semaglutide, but that's not likely to be the situation. So I think we will refrain from further comments, Peter, on that one until we present the data. I think that's the best thing to do. When it comes to stem cells, we've been active since the 1990s and we're actually very active and that means that we also have gone down the differentiation path from human embryonic stem cells all the way into something that is active, in-vivo glucose-sensing adult phenotype cells that do all of the right kind of things. We're just less publication active because we're pursuing it more as a therapy than as an academic target. There are many barriers still to overcome in terms of encapsulation, oxygenation and a multitude of other things to think about. But it's very exciting and there is progress both at Harvard and at Novo Nordisk and elsewhere. When it comes to--

Maybe just to add what is the magnitude of the investment, because I also heard that you had that element in your question. And I would anticipate that this would only cover to the tune of 1% to 2% of the overall research investment in diabetes care. So it's a small, exploratory area. It's not a major investment area.

No. And to be honest it's so sophisticated that this is something that for those people primarily that have brittle diabetes that have very volatile glucose excursions despite optimal insulin therapy. So it's not a treatment for the masses, so to speak, it's very tailor made also going forward I would expect. When it comes to Alzheimer, this is a high-risk area, as you know. There is already been ongoing a small pilot study under Danish academic conditions looking at GLP-1 therapy on Alzheimer's and there's a bigger one ongoing in early Alzheimer's patients that still have a reasonable cognitive function. And that's ongoing, driven by the Alzheimer's Society in UK Imperial College and co-funded by Novo Nordisk. It's a phase 2 study in 200 patients. It's still recruiting. And it is exciting. There's a lot of interest in the whole role of insulin signaling in the Alzheimer's, but whether this has to do with insulin signaling or GLP-1 signaling or signaling at all, time will show.

This is Lars Rebien, just a final comment on stem cells. When we look at it, as Mads rightly pointed out, it is likely to initially be a relatively modest number of individuals that will benefit from this. For us, getting into this field will be quite costly, eventually, if we can convince ourselves that we have something which can maybe change, because it's different types of lab that are required, different types of manufacturing which is required, encapsulation and all the rest of it. So in fact, when we look at it from a financial perspective it's very difficult for us to really make a very good financial case. The reason why we're being involved is that we want to change diabetes and we believe this is one of the routes and so we will do it regardless of whether it's a huge financial upside just for the sheer sake of doing it.

We will now take the next question from Richard Vosser from JPMorgan. Please go ahead.

Two on the GLP-1s and CV trials, please. So given what you know from the results of LEADER and now SUSTAIN 6 how would you progress about designing further landmark trials for semaglutide? Would you go into earlier diabetic patients or do you think you've got the sweet spot at around a 3% event rate in patients? And then secondly, obviously, we've got two significant CV benefits. Given the relative size of the benefits that you've seen do you see the potency of the GLP-1 molecules playing a large role in the magnitude of the CV benefits? Anything you could give us there would be very useful. Thanks very much.

Yes. I think let me start by saying one thing that was not so clear in the brief company announcement yesterday and that is that the populations we're treating in SUSTAIN 6 and LEADER respectively are very similar. And since you know the baseline demographic paper is published for LEADER a long time ago you can actually look at what are the patients and they are as I just told you half an hour ago. So it is true that these are relatively longstanding people - diabetics with a relative mature age 64 and a decent BMI of around 33. And you can argue, as I had done historically, that moving into younger age should enhance the [indiscernible] ratio. But when that is said we also have to admit that this is a rather ill population and yet we were able to see the clearly significant effects in both of these trials which is a testimony to the mechanism of action, whatever it may be. It's probably several mechanisms of action coming into play, actually, when we're talking our two molecules - is something that seems to lend itself even to more longstanding diabetes which is a little bit contrasting what I would have told you some time ago. There are some signs that maybe younger age could be a predictor of better response, but these are things that will be taken duly into consideration as we design a, quote unquote, landmark study or a bigger study, for instance, on semaglutide going forward. Then when it comes to the differences between these molecules liraglutide and semaglutide only differ by one amino acid, namely, position eight and then they differ by the nature of the linker and the nature of the side chain. Other than that they are relatively similar molecules. Yet there are rather striking pharmacological differences in terms both of potency and apparently also efficacy. And whether that plays out in the cardiovascular outcomes we will have to wait and see when we announce the full data sets.

This is Lars Rebien here, just an additional comment on how to progress with sema. It is clear if we get visibility on the ability to deliver sema as a tablet that would, of course, beg us to consider testing it over long periods of time in an earlier population to show the benefit and whether that will be a cardiovascular benefit or other benefit, preferably deferring late complications in people with somewhat earlier type-2 diabetes.

Next question comes from Martin Parkhoi from Danske Bank. Please go ahead.

Two questions and I'll stick to them. Firstly, on the Tresiba Levemir split you showed the market share on new brand and from that we can see that around 60% of the Tresiba share has been taken from Levemir which is higher than you have communicated before. Is this a thing we should see in the beginning and expect - do you expect going forward that we should see higher gain from the glargine franchise? And then the second question this is regarding Canada. We have seen one of the first markets in Canada what the treatment guidelines actually have been changed following the CV data on Jardiance. I know it's very early, but can you say anything of if it has changed anything and do you also expect that there will be a change to GLP-1s in general when the detailed data has been published on the ADA in June?

Mads, what are your anticipation? It is clear that we in the beginning have cannibalized a bit more Levemir patients than we had anticipated ourselves. So basically the origin is one-third from new starts, one-third from Levemir and one-third from glargine. What are your anticipates longer term, Mads, in terms of the nature of the patients?

I think very often what happens immediately is, obviously, that there is a trend towards those physicians that have confidence in your historic products are maybe more likely initially to switch patients from the previous generation to the new one. But when that is said, then, once you gain experience with a product you know what it can do and in which patients it serves its job best. I think it's more likely, a, that you will see a trend towards more patients not being switch patients, but new-to-insulin patients and that trend in a way, I guess, is already a little bit ongoing. But this is often what you see in - yes, in the early phase of a launch. And then there is a specific thing about Levemir and whether this is true or not [indiscernible] on the short side of action for once daily in some patients. And since Tresiba on the one hand offer a unique long half-life and also is available in the U200 formulation there are of course some patients that lend themselves well to another switch. But I think over time the game is about, of course, new-to-insulin patients rather than switching, but this is the early launch phase.

Before I hand, Jakob, the question on Canada and whether the changed guidelines have had any implications for GLP-1s and what we could expect going forward, perhaps, Jakob, you could share your perspective on Japan, where the origin of the patients are coming from in Japan, where we have had Tresiba on the market for quite a bit of time and that we now have one-third of the basal market in Japan, whether we can make any correlation between the evolution in Japan and over to the U.S.

No, Japan is clearly more patients that are new to basal, because we're further progressed in the launch. And also seeing a decline in the proportion that comes out of Levemir that's also a pool that's getting smaller and I would expect the U.S. situation to transition towards something that was similar.

So that corresponds to this-?

Yes. So it's a difference in what we expected, but it's something we'll grow out of and over time will solve itself.

And then Canada?

Yes, Canada, I cannot make a very quick comment on guidelines, because often one believes that if you make guidelines then physician behavior will change. But often I think one should see guideline as a result of physician convictions. So the fact that the Canadian guidelines have been changed is a result of the Canadian key opinion leaders supporting the relevance of cardiovascular data for the SGLT-2 class. So what we see in terms of growth of SGLT-2 is in my book more a result of key opinion leader attitudes that are very positive than the guideline per se. And of course, then it becomes an open question what will happen when we get data in the GLP-1 class. And the expectation is that in different places around the world there will be similar discussions on whether that shouldn't cause a change in the treatment guidelines. But I think the fundamental dynamic will be when the physicians, the key opinion leaders, are getting exposed to the data and with their own eyes see the impact of that that will start changing behaviors.

Next question comes from Michael Novod from Nordea. Please go ahead.

Two questions, one on Levemir. And despite actually seeing some switching into Tresiba you still have 11% growth local currency in the U.S. Could you talk a bit about that? And also then related a bit to the European pressures, 3% decline, is that due to Basaglar? Or is it due to you, say, adopting a lower price on Tresiba so we see stronger switching? And then the second thing also relating to semaglutide. Going even, say, into other indications we've seen Lilly starting now or announcing Jardiance to be tested in chronic heart failure. Is that also a road you could decide to take if you see evidence for semaglutide orally being successful?

It is correct that we see still quite strong growth of Levemir in the U.S. There is some volume, but there is also a price effect. We took a price increase last year, so I would say probably half is price and half is volume of the 11%. In Europe it's basically cannibalization from our launches of Tresiba that account for a decline in the Levemir segment, but also a reflection of the competitive environment where we now are seeing Basaglar and biddings in Europe being quite aggressive. Mads Krogsgaard, do you have - in addition to what everyone else might think of NASH, whatever have you, chronic heart failure, will you expect to go into this?

Yes. So obviously when you have a relatively acting hormone like GLP-1 you have to think about which kind of biologies and hepa-physiologies you can target. And the one that begged itself for the beginning was type-2 diabetes. This has been followed by obesity and lately we're moving into NASH, as we've communicated. And then, driven by the LEADER and SUSTAIN 6 trials it's quite evident that Novo Nordisk will continue with an emphasis on semaglutide, future investigations into the microvascular area, i.e., cardiovascular. And whether we will split that in classic anti-atherogenic effects, pursuing myocardial infarction prevention, stroke and death or more into the area of congestive heart failure which is a slightly diverging but related area, is something that we'll communicate much more about as we move forward, as we also will, by the way on the potential for GLP-1 that has been speculated also to improve microvascular performance over time or at least prevent microvascular disturbances. So all of these things I would look very much forward to discussing with you perhaps later this year.

Our next question comes from [indiscernible] from Credit Suisse. Please go ahead.

First one is the benefit seen in SUSTAIN 6, was that across multiple elements of MACE? We know Jardiance had a benefit only on the heart failure side and you mentioned in the talk that Victoza's LEADER trial had the reduction driven by mortality, MI and stroke. That's my first question. And secondly you mentioned in 4Q we potentially may see the approval of your ultra fast-acting insulin. How much of a need is there for an ultra fast-acting insulin and you think payers will be willing to pay a premium for this type of drug? Thanks very much.

Mads, have you seen benefits in all the three segments of the strict MACE or as we did see with LEADER? Or how would you characterize this?

Yes, so two things. One, this is by definition a much smaller study with only 250 MACE events versus more than 1,200 MACE events. That means that it is a small-numbers game, but I would reveal so much as to say that there are none of the three endpoints in the strict MACE being mortality, stroke and myocardial function that is at or exceeding unity, being 1.0, that far I will go, but not give you a split until we reveal the data at our conference, hopefully, second half. And then should I give a snap view on FIAsp-?

Yes, you can do that and then we can ask Jakob to comment on what he expects from a commercial perspective in terms of positioning this.

Clinically - that will be my comment. Clinically, the thing that we have seen, in particular, in type-1 diabetes with a fast-acting aspart - well, actually, it's three things. It seems to be a very suitable pump insulin based on one study. We're investigating that much more going forward, where there is clearly a need for a better pump insulin. FIAsp looks to be such an insulin. When it comes to dosing flexibility we have shown that you can post-meal dose and yet achieve the same outcomes which, for many patients, can be a good factor, including both kids and elderly, but also erratic eaters, so to speak. And then the other thing is that there is actually an HbA1c benefit, albeit modest. So it's not a revolution, but it is an evolution, we believe, in prandial insulin therapy. And the differences between this one and NovoLog/Humalog is basically of almost the same or more or less the same magnitude as when the analogs came of age compared to human insulin in the old days.

So, Jakob, what are your expectations from a commercial perspective?

Yes, how we get the payers to pay for this benefit. Let me comment. One additional comment to what Mads said is we also see a lot of other developments in technologies that will involve continuous glucose monitoring that automatically through algorithms will guide patients. I think that's a very promising area that will help a lot of patients. And when you operate with that anything you can do to minimize the time gap between administration and action is going to greatly improve the accuracy of these algorithms. So in that sense, yes, there is definitely a need out there. In a broader type-2 population one could, on the one hand, argue is that going to be seen as a big difference when you administer faster aspart. I think one has to note that you're not asking patients to do anything different. They're going to have an easier time administered and it is going to cause a reduction in hypos. That's always worth pursuing. And then last to you your final question, okay, would that be able to command a premium in the market? And of course you can say once we're through, we see the final label, we see what claims we can make in the market we'll make a decision on the premium. But it's expected to be modest and we would probably prefer to get the technology introduced broadly rather than emphasizing too much a premium in this case.

Next question comes from Tim Race from Deutsche Bank. Please go ahead.

In the interests of sounding a little creepy, well done on the semaglutide unexpected positive readout. But just on that element, in terms of GLP-1s, could you just help us think about the patient share that GLP-1s have today in terms of absolute patients rather than value share and where you actually might be able to get to given the positive data we've seen from the class at the moment? And also perhaps could you just help us understand what the average stay time on a GLP-1 therapy is today based on its ability to control HbA1c, so how long do patients take GLP-1s for on average generally? And then if I may just a question regarding biosimilars in Europe and Japan. Could you just talk about the extremes that you've seen in different markets, of which markets you've seen very little impact and why and which markets you're seeing more of an impact and why?

I'll try to deal with the first question on GLP-1, where we're right now in terms of value and what it might look like. And then I could ask Jakob to talk about his view of the average stay time and also comment on the impact on biosimilars, because I think the most pronounced impact for biosimilars is actually in Japan for some regulatory reasons. But if we look at it - we said when we introduced Victoza that it was our anticipation that we would be able to build the value of GLP-1s then to 10% of the diabetes market. This is where we're approximately now in United States. There are of course lots of markets in the emerging markets where GLP-1s for cost reasons are - have not been introduced. But given, then, the data that we have seen both on LEADER and on semaglutide, given the possibility for Novo Nordisk to in the future have a portfolio of drugs, one of which, Victoza, will eventually run off patent while we still have patent protection for semaglutide as a once weekly and eventually an oral entry, I would expect us to be able to in the future penetrate significantly more in emerging markets with using different price points than we can today. And so I'm of the opinion that I see a much larger potential for GLP-1s. And I would not be surprised if we end up with 20% of the total diabetes market being accounted for by GLP-1s in a 5- to 10-year horizon. Jakob, current understanding of stay time?

Yes, it's a question I think we started receiving right when we launched Victoza and it's oddly enough a very, very difficult question to answer. So what we know is that after a year somewhere between 50% to 75% of the patients are still on Victoza in crude numbers. But there are a lot of factors influencing this beyond, you could say, the glycemic control you are achieving. There are changes of plans and generally we see a lot of shifts in the pharma market that are not necessarily driven by clinical outcomes. And then you could say as we look ahead we come back to the CV data. Once there is this understanding that there are implications of GLP-1 treatment or longer term implications from a cardiovascular protection, is, all other things equal, probably likely going to give a motivation to stay on the product for longer time. But what we see is it's slightly shorter right now for GLP-1s versus GLP-2s, on par with DPP4s. But again, I said, it's hard to derive any conclusions. It's hard to derive any hard numbers on that.

So biosimilars impact, I know in Europe it's relatively more isn't it? It's certain countries there for specific reasons have pricing regimens which favors biosimilars where we've seen an impact. What about Japan?

Japan, what we've seen is a fairly big dent into the Lantus franchise from biosimilar glargine, but at the same time we've seen an uninterrupted continuous penetration of Tresiba. So it confirms what we also discussed before. So far what we see is that it's within the molecule and there are no wider implications of biosimilar competition beyond the actual molecule it is substituting.

We will now take the last question from Ronny Gal from Bernstein. Please go ahead.

Two of them, first, biopharmaceutical investments, so we've seen some of your peers reconsidering your investment in this field. Your thought? Any change in your view of the biopharmaceutical markets? And second, Express Scripts, now with an [indiscernible] of new data about cardiovascular benefits in GLP-1 and in Tresiba. Is there potentially room to be re-introduced into that large payer or this is something we should not be holding our breath for?

I'll take or at least I'll attempt to answer the questions. Biopharm, it is obvious that when we look at the biopharm contribution to growth it is less than the diabetes business for us. We have also heard that there are rumors in the market that other of our peers are considering their investment in the field. We're not considering our investment in the field. We think we have a strong franchise. We think there are good clinical reasons why the NovoSeven business is likely to prevail also, even though new technologies may be introduced. We have an exciting pipeline and we're putting maximum force behind ensuring that that pipeline is further developed. We're expanding manufacturing capacity to be able to launch as quickly as possible our best-in-class factor IX GP. And as you know we also have an VIII GP, a long-acting factor VIII. So we're pretty satisfied with the biopharm business. We also have advanced our long-acting growth hormone which is part of our biopharm franchise. So we have no intentions of divesting our business, so that's that. ESI, it is of course something which is catching interest amongst the payers that wants to provide their new and innovative drugs to their customers that are buying their insurances. So I'm sure that the recent results on Victoza and the switch data on Tresiba is going to help us in the discussion with ESI going forward. I guess that's as far as I can get during this public call, but we're quite excited about that opportunity.

So with that, ladies and gentlemen, I would like to thank you all for being all of you almost limiting yourselves to two questions, but we're in a good mood this afternoon. So we thank all of you. And we're looking forward to return and see some of you during the ADA and then return with the half-year results and then EASD later in the year. So thank you very much for listening and if you have any questions please contact our investor relations folks. They're sitting by their phones. Thank you.