Good day, and welcome to Q1 2013 Novo Nordisk A/S Earnings Conference Call. Today's conference is being recorded. And at this time, I'd like to turn the conference over to the CEO, Lars Rebien Sorensen. Please go ahead, sir. Lars Rebien Sørensen: Thank you very much and welcome to this Novo Nordisk conference call regarding our performance in the first 3 months of 2013 and a new 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 with us are our investor relations officers. And today's earnings release and the slides that we'll be using for the call are available on our web page, novonordisk.com. The conference call is scheduled to last approximately 1 hour and, as usual, we'll start with the presentation as outlined on Slide #2. And the Q&A session will begin in about 25 minutes. Turn to Slide #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 results to differ materially from expectations. For further information on the risk factors, please see the earnings release and the slides prepared for this presentation. Please note that this conference call is being webcast live and a replay will be made available on our Novo Nordisk website after the call. Turn to Slide #4. We're pleased with the strong sales in the first 3 months of 2013. Sales increased 14% in local currencies and 13% in Danish kroner compared to the same period in 2012. Sales growth was driven by strong performance in North America, International Operations and China. Sales growth was realized within both diabetes care…

Thank you, Lars. Please turn to Slide #15. The regulatory reviews of Tresiba and Ryzodeg continue around the world. New drug applications under review in 18 countries and approvals have been obtained in a number of markets in addition to the 3 countries, which Tresiba has already been launched. Since we announced our full year results, approvals for Tresiba and Ryzodeg have, thus, been obtained in Norway and Iceland, and approval for Tresiba has been obtained in Switzerland. Please turn to the next slide. As previously announced, we received a Complete Response Letter from the U.S. FDA regarding the new drug applications for Tresiba and Ryzodeg in February of this year. In the letter, the FDA requested additional data from a dedicated cardiovascular outcomes trial before the review of the new drug applications can be complete. We are currently in a dialog with the FDA regarding the detailed design of the cardiovascular outcomes trial to generate the requested data needed for resubmission. The trial is expected to be double-blind, event driven, based on the needed number of cardiovascular events with insulin glargine as comparator on top of standard of care. It is expected that the basis for submission will be an interim analysis demonstrating that the upper bound of the 2-sided 95% confidence interval for the estimated cardiovascular risk ratio is lower than 1.8. With the reassuring point estimate for the hazard ratio. In addition, it is expected that we will be required to continue the trial to completion after approval in order to rule out the relative risk of 1.3 based on the 95% confidence interval. We expect to initiate the trial within the next year and that data for the interim analysis will be available 2 to 3 years after trial initiation based on the underlying assumptions in…

Thank you, Mads. Please turn to Slide 17. In the first 3 months of 2013, sales increased by 13% measured in Danish kroner to DKK 20 billion and by 14% in local currencies. Sales growth was positively impacted by a number of nonrecurring events, including timing and sequence in the U.S. and International Operations, extraordinary sales in International Operations, as well as a modest level of sales from NovoSeven in the first quarter of 2012. Reported gross margin improved by 110 basis points to a 81.9% in Q1 2013, primarily driven by a favorable price development in North America and the positive net impact from product mix due to increased sales in modern insulins and Victoza. The gross margin was positively impacted by currencies by around 10 basis points. Total non-production-related costs increased by 11% in both local currencies and in Danish kroner. S&D cost increased by 15% in local currencies and by 14% in Danish kroner to DKK 5.5 billion. The growth in cost is driven by the expansion of the U.S. sales force in the second half of 2012, costs related to the launch of Tresiba in Europe and Japan, as well as an impact from the reversals of the equal provisions in the third quarter of 2012. R&D cost increased by 6% in both local currencies and Danish kroner to DKK 2.7 billion. The relative modest cost increase is reflecting timing of clinical trial activities and is primarily driven by development cost related to the ongoing Phase IIIa trials for the once weekly GLP-1 analog, semaglutide and liraglutide in [indiscernible] . Within biopharmaceuticals, costs are primarily related to the continued progress of the portfolio of development products -- projects within hemophilia and the Phase II trial for Anti-IL-20, a recombinant human monoclonal antibody in rheumatoid arthritis. Operating…

[Operator Instructions] And we'll take our first question from Richard Vosser from JPMorgan. Richard Vosser - JP Morgan Chase & Co, Research Division: It's Richard Vosser from JPMorgan. One question on Tresiba, please. You mentioned, I think, that the hazard ratio needs to be reassuring at the interim analysis. Could you explain what your discussions with the FDA are on that point, whether the hazard ratio needs to be below 1 and, I suppose, your confidence in that given the previous data had a -- has a ratio of 1.1? And any update or any more details you can give when you'll be able to finalize the protocol with the FDA would be useful, too. And then just second question, just on Victoza sales. Just perspective on Victoza sales development, how you see that progressing with the incremental competition this year and any impact on the sales from pancreatic cancer concerns that seems to have blown up recently? Lars Rebien Sørensen: Thank you very much, Richard. This is Lars here. Mads, I'd like to defer the first question on Tresiba and the discussion around what an reassuring endpoint is at the interim stage. And then I'll come back with a comment on Victoza.

Yes, well, first of all, Richard, in principle, one doesn't comment in great detail on an ongoing dialogue with the regulatory agency. But what you should assume is that we are -- you can argue, more or less, in the situation of a classical diabetes drug for Type II diabetes that you are ruling out, one, you can say upper bound at the interim that is suited for approval and then, another one, namely 1.3, for the post-approval situation. And as always, you'd like to have a reassuring point estimate. And one way of looking at the reassuring point estimate is one that when you approve a product based on that, then with a reasonable degree of certainty and statistical probability, will actually enable the company to, as they move to the completion of the trial, i.e., when they have put together the total amount of [indiscernible] in the trial, will be able to also rule out the upper bound of 1.3 at completion. So you can actually use a probability statistics to evaluate what a reassuring point estimate is without going into much more detail. Lars Rebien Sørensen: Thank you very much, Mads. And then with regards to Victoza. We can give you the guidance, I think, which is very consistent with what we have been giving you in previous quarters. The continued strong growth of Victoza in an absolute terms, this would be around DKK 150 million to DKK 250 million added in sales per quarter. Please use Mad's data if you want to make projections because individual quarters tend to vary. In regards to any impact of the public debate on pancreatic safety, we have not seen any impact so far. It is a public issue, which has been raised and also litigations have started in United States as we all expected. And but so far, no impact on the penetration, and hearings are going to be held in the offices of NIH in June. Dealing with the pancreas where -- in a sort of a more neutral setting, scientists will be presenting their pros and cons and their scientific data to validate their position. And hopefully, in our view, this will clarify issues more than the rather, let's say, populistic media attention it has gotten so far. So we cannot give you any certain guidance other than it has not had any impact so far.

We'll take our next question from Michael Novod from Nordea.

It's Michael Novod from Nordea Markets in Copenhagen. First question to Victoza in Japan. Maybe you could also provide a bit more flavor on the situation because you can see 0% growth in local currency in Japan. I guess, it could be nice to see a bit stronger development. Any intentions of trying to, say, boost that going forward? And then secondly, regarding the Tresiba trial, CV trial, you're talking about -- or you say it's going to be a double-blind. How are you going to do that? Would that be without a device, then? Would it be in a vial and syringe or how should we see this being conducted? Lars Rebien Sørensen: Thank you very much, Michael. Lars Rebien here. Mads, if I may first go take a shot at the Victoza Japan, then you can deal with expectations for the design of the trial for Tresiba. With regard to Japan, our market situation in Japan is not as optimal as we would have liked it. The doses of Victoza in Japan approved was only 0.9 milligrams, which is not, in our view, an optimal dosing. However, this was what our clinical trials actually had to deal with because we followed some recommendations from the Ministry of Health and Welfare in Japan when we designed these trials. Then I will remind you that we had some very unfortunate incidents in the beginning of the launch of Victoza in Japan where we had a couple of fatalities because physicians, misunderstandingly, took insulin-demanding patients and put them all on insulin and put them on Victoza, which, of course, is not acceptable when they are type 1 patients or insulin-dependent patients. So all of this together has meant that the penetration of Victoza in Japan has been less than our expectations and we have no real expectations that this is going to change dramatically because that would require new data and a new dosing regimen. So we're a little bit on the back foot with Victoza in Japan. Of course, in the future, IDegLira poses a tremendous opportunity in that we have Tresiba approved. We are gaining market recognition of Tresiba already in the first weeks, and the combination of Tresiba and Victoza would be an ideal product, in our view, for Japanese diabetics. And hence, we are optimistic that would change the game in Japan. Mads, how are you going to go about making a double blinded study with glargine and Tresiba?

Well, Michael, as you correctly allude to, it is so that double blinding will entail the use of vial and syringe. And just to make -- also bear in mind that when we talk about insulin sales to worldwide and in the IMS market, and so on, we are typically talking about values and value markets and so on and so forth. In reality, if you take the amount in emerging markets, but even in the United States, our patients who are using vial and syringe, this is a humungous amount of insulin users. So yes, it will be vial and syringe based. That's the way to totally blind it, but that is actually also doable. There will be some markets where it's -- some territories where it's less doable. That's a fact. But obviously, as we always do, we entertain the feasibility studies into which markets, which investigators we can get the respectable recruitment times that we normally do in our company.

And we'll take our next question from Peter Verdult from Morgan Stanley.

Peter Verdult here from Morgan Stanley. Just one for Lars, Mads and Jesper, if I can. Any change or update in terms of expected impact from U.S. global healthcare reform this year, and maybe a quick word on any incremental measures that you're fearing or expecting in the U.S., specifically? A quick one on Mads for semaglutide. I see you've gone for the 1 milligram top dose into Phase III. And if I remember correctly, in Phase II, it was only the 1.6 milligram dose that actually showed superior efficacy to liraglutide, but, obviously, with a pretty nasty side effect profile. So I just wanted to get an understanding, with this 1 milligram, are you still hoping to show superiority versus liraglutide from an efficacy basis or is this just sort of the once weekly liraglutide? Some update there will be helpful. Then a quick one for Jesper, just can you quantify the legal provision release in Q1 and just give us an update on the infrastructure you currently have in place in Europe. I mean, is there scope for that to be reduced or the need to maintain to support Tresiba launch? Lars Rebien Sørensen: Thank you very much. That was a nice one, one for each. Let me start out with the -- our expectations for health care reforms, and particularly in the United States. There's still going to be some annualization of the implementation of the remaining parts of the healthcare reform in the United States. So far, our guidance is not including any additional measures. We take notice of the administration's proposal, which had some proposals in them which are quite onerous -- and would be quite onerous to the industry in general, but also to us. I'm thinking here in particular about the dual eligibles, but also measures such as covering a larger part of the donut hole. But we have not taken those in our guidance for 2013 in as much as we do not expect that they will have any impact this year. Mads, a little bit on semaglutide, the dose selection and considerations for why you've chosen the dose.

Yes. Well, first of all, Peter, one comment on the side effect profile that I do know that you're aware of, that is the notion that we kind of underestimated the potency of this agent in Phase II. This means that, actually, we went directly onto doses that were, you can see, reminiscent of 1.8 or more milligram of Victoza already in the first go, i.e., the 0.8-milligram dose without titration. So there's no doubt that Phase III is done with a much more subtle titration algorithm, and it is the subtlety with which you titrate any GLP-1 agonist that actually dictates which degree of GI disturbances you will envision. So I think we're in control of that part. If you look at the pharmacomathematical or kinetic modeling of those Phase II data, it actually shows that you will come to the conclusion that 0.4 milligram witness more or less the same degree of glycemic and weight control as the 1.2 milligrams Victoza open arm that was included in that dose range finding study. So the one 0.5-milligram dose that we presented for Phase III is quite clearly the equivalent of 1.2-milligram Victoza. And the 0.8 milligram actually did slightly, but not significantly, at least not on old parameters, better than the 1.8-milligram Victoza. However, we are not using 0.8 milligram, we're using 1 milligrams, and the modeling that we had done will actually dictate that the 1-milligram dose will be the most efficacious agent that we know of that is into Phase III. But, obviously, all of this is modeling. We expect a strong efficacy both in weight and in glycemic parameters for the 1 milligram and now we're looking forward to seeing the data. Lars Rebien Sørensen: So thank you very much, Mads, for that elaboration. And Jesper, legal variations release in the first quarter and further comments to that?

Yes. It was the release of legal provisions in the first quarter of 2012 and this release of legal variation has a net effect of about DKK 150 million in the numbers when you compare them. In terms of sales force in Europe and our actual roll off of Tresiba, we believe we have an adequate size of sales force and we will continue to deploy it effectively with both the rollout of Tresiba, but also with the continued stated penetration of Victoza in the European markets.

And we'll take our next question from Martin Parkhoi from Danske Bank. Martin Parkhøi - Danske Bank Markets, Research Division: Martin Parkhoi for Danske Bank. I have a question regarding your long-term targets. You set out some new targets in connection with your full year results. But after that, of course, Tresiba, Ryzodeg and IDegLira has been delayed in [indiscernible]. You do not comment on these long-term targets in connection with Q1. Is it because you still feel confident with these targets of a 15% EBIT growth or is it because you only make adjustments to them in connection with full year results. And then a little bit related to this, given that you now have to put more marketing effort into your, you say, older products. I think I've got the impression historically that if you can grow your top line by 10%, then you can also grow EBIT by 15%, but will it be more expensive for you going forward to grow your topline by 10%? Lars Rebien Sørensen: Thank you very much, Martin. This is Lars Rebien here. I'll just give a few comments, and then I'll ask Jesper to add if there's anything that I've left out. You're correct, we did retain the newly issued long-term guidance when we came up with the announcement that we had a Complete Response Letter on Tresiba. And I think we did comment that it was our intention, when we came out and gave you some guidance for '14, that then we would validate that those targets, indeed, were still realistic. With regard to the overall model, it is correct that we on -- typically on a top line growth of 10%, 10% plus, should be able to generate an operating profit growth of 15% given the mix changes and given increased productivity and, also, leverage it down through the P&L because we now have a global organization and it is obvious it's -- the market, towards the end of the patent runout of Lantus, will become more and more competitive. And so the marginal return from investments in sales and marketing is, of course, declining. So we do not exclude that it will be harder effort. However, I would say that we have been positively -- not surprised, but we've been pleased with the development we've seen in recent quarters in the United States of our sales and marketing effort and the gain in market share that we've had. So we intend to continue with that and then we'll be able to give you a midterm update on the long-term targets when we see how we fare for the full year. Jesper, anything you want...

Just on the F&D ratio, I think we'll be in the vicinity of 28%. And what I think you are rightly alluding to in terms of the additional marketing investment required near term to maintain the top line growth, that will, everything else equal, keep us closer to the current level of around 28% in F&D ratio. And an improvement in that ratio is probably not as likely in '13 and '14, '15, as may have otherwise been the case. And it's just really a case of us redirecting a bit of investments towards the long-term, high-growth potential market in International Operations and, thereby, maintaining the F&D ratio. We'll continue to -- we hope to make improvements in our gross margin development and we'll continue to make efficiencies in terms of revenue cost, and through that, hopefully, and gradually, in total, still be able to expand our operating margin.

And we'll take our next question from Peter Ankersen from ABG.

Peter Ankersen of ABG. Firstly on some of your competitive products in the DPP-4 market, it seems as if the market has kind of matured. We just saw Merck reporting a decline year-on-year. You previously used the DPP-4, the incretin mimetics, as an explanation for the lack of growth. How do you see the situation now, and do you really think that this maturation of the market is real? And then secondly, maybe, Mads, could you give us some more flavor on the CV study in terms of size of -- number of patients and also in terms of the event rate you expect for the interim analysis. Lars Rebien Sørensen: Thank you. This is Lars Rebien here. I'll just give you a comment on the DPP-4 market. We do believe that the slowing down or the announced declining growth rates in the DPP-4 market still leaves an opportunity for the GLP-1 market to grow. We're not seeing any slowdown in penetration of the GLP-1, the GLP-1 agonist, that is. So we still think that we will have an opportunity to continue to grow Victoza, as I've already outlined. And more interestingly, perhaps, is that we do believe that we have felt an impact on the insulin market from the penetration of DPP-4s, and, hence, if your assumption is correct that the DPP-4s are not growing that much anymore, then we might see the beginning or the resumption of growth of the insulin market, which will be beneficial for Novo Nordisk. So that would be my comment. In general, we need to realize that we talk -- that Merck and others talk about the increase in class, which is a misnomer. We're talking about 2 agonists, exenatide, which is very similar to human GLP-1; Victoza, which is human GLP-1 analog and then the DPP-4s, which are completely different types of molecules that act completely differently pharmacologically and that is something to bear in mind. Jesper -- or Mads, you get the last one.

Well, Jesper can comment on the CV study, if you want.

Okay. Now the hint I can give you, Peter, is that in as much the FDA never issued specific statistical analysis considerations in their formal guidelines for CV outcome studies, they did make some deliberations in some associated slide presentations that they've made on different occasions. And in those, they talked typically about the statistical power to rule out 1.8 and 1.3 of 90%, respectively, with MACE events, to the tune of 122 for 1.8 interim and 620 or so, or maybe slightly less, for the 1.3. Now so you can imagine that if Novo Nordisk wants to have a reasonable degree of certainty, we would, if anything, shy -- or go on, on the side of more power to our analysis such that we are sure to get it -- as sure as possible to get a good result for this insulin. And then in terms of the event rate and the amount of patients. Well, the amount of patients -- the LEADER trial is about 9,000 patients. You can imagine that these patients are probably more into the same kind of inclusion criteria as we've used in the LEADER. But since they have progressed more down the treatment [indiscernible] of diabetes in that they are now insulin demanding, or needing, then, basically, if anything, the event rate would be slightly higher and that would mean an invalid that will be slightly lower, but I would shy away from commenting specifically on the amount of patients in the trial, but these are some considerations. I think, event rates, typically nowadays, in this kind of population, do tend to be on high side of 2 -- well, below 2, but they also rarely go above 3. So these are the ballparks we're talking about.

We'll take our next question from Tim Race from Deutsche Bank.

First of all, just on -- from Mads on obesity, the 1.8-milligram versus 3-milligram dosing that we saw in the diabetes study, how confident are you that you've got the -- for the bulk of the trial program, you've got slightly the lowest effective dose. Like you said in Japan, for instance, regulators can chose lower doses, et cetera. What have your discussions been with the regulators designing this trial program, given that the obesity and pre-diabetes study is only at 3 milligrams? And then just moving onto the Tresiba CV study. Will you also be collecting hypoglycemic data? I mean, I know it's a second level point, but would you be able to finally put to bed your benefits claims versus Lantus through this trial? Lars Rebien Sørensen: Thank you very much. These are very good questions. Mads, you're on it again.

Well, indeed, a good question. So I'm happy that we included the 1.8-milligram dose in this particular trial, the SCALE diabetes study. And why I am happy? I'm happy because not only on glycemia and weight, but also on a number of secondary supportive points, such as cardiovascular risk and so on and so forth. That we see a statistically significant separation out of the 3-milligram dose favorable to that of the 1.8 milligrams, such that we had a dose differential that proves the point that there's a meaningful difference between the 2 doses. On the other hand, I'm also comfortable that if we can hit 1 of the 2 FDA criteria in the diabetic population with that dose, it should actually be easier to hit both of those points in the big SCALE obesity trial that's coming later this quarter. But, of course, we will have to deliver those data and then we'll see what they show. So I actually think that if you talk about a minimally effective dose, you, on the one hand, need something that prove those at least 1 of the 2 FDA criteria, whilst at the same time, differentiating itself preferably against existing doses that you have on the market being 1.8 milligrams. So you could have thought about the 2.4 milligrams dose and we don't have data, that a part of in Phase II. But in Phase II, the 2.4-milligram dose, I would say was borderline against the FDA criteria, whereas the 3 milligrams was robustly above. So I think we've given that due consideration, actually, but time, hopefully, will prove me right and not wrong in that regard. When it comes to the CV study, we had tried to make it as simple as possible, but you are mandated to always collect severe adverse event [indiscernible], and one [indiscernible] is of course, severe hypoglycemia. So since the trial is large and long, obviously, you will have a gathering together of severe hypo events and, of course, that can solidify our story, but Novo Nordisk may also be doing other things to further solidify the hypoglycemic case over and above what has proven prudent in Europe and Japan, but also something that will support the U.S. case at that point. Lars Rebien Sørensen: Thank you very much, Mads. Yes, it is indeed the case that we have a little bit of a dispute also on understanding the hypo benefits with the FDA, with use of Tresiba and is, of course, in our interest then to see if we can further our argument in that regard before we get to a point where we need to resubmit in the U.S. So we intend to do so. How we intend to do so? We'll be disclosing later, perhaps. It's a little bit technical and complicated at this point of time.

And we'll take our next question from Carsten Madsen from Carnegie Bank. Carsten Lønborg Madsen - Carnegie Investment Bank AB, Research Division: Carsten Madsen, Carnegie, here. Just one more question to the CV trial. I was wondering if you can shed some light on which type of MACE you'll be looking at. I guess we were a little bit surprised when we saw the briefing documents. We had MACE, we had MACE plus, we had your analysis, we had FDA's analysis. So I do know if you can tell what type of MACE, specifically, MACE or MACE plus, you will be looking at in the CV trial? Lars Rebien Sørensen: Thank you very much. Mads?

Yes, well, Carsten, if you go back all the way to 2009-ish, where we had the Advisory Committee on Victoza. You'll recall that there was so many different definitions of MACEs introduced by regulators, such that we could come up with probably a high-teen number of different analyses for Victoza. Fortunately, giving us, I would say, a robust situation, which, I guess, was also the reason for fasting for these analyses. Now there's also been a debate both on obesity outcomes and Tresiba outcomes and others about whether to use a 4-string or a 3-string or some other MACE. And I think the FDA is leaning towards small-scale pre-approval studies such as these IIIa trials, typically accepting what you can call MACE plus, which is a MACE where you also include things such as unstable angina, hospitalization requiring emergency treatment, so on and so forth. Whereas, they tend to lean towards having what they call strict MACE, or the 3-string MACE, constituted of -- or comprised of cardiovascular fit [ph], nonfatal MIs and nonfatal strokes. So I think you should imagine that this is more to the kind of large-scale kind of trial where you're going for the real hard end point, which is the more strict MACE with the 3 elements. Lars Rebien Sørensen: Thank you, Mads. This is Lars here. But you should remember that at the end of day, the data is the data, and the agency, if they find something in the soft analysis, they will still try to hold response our account [indiscernible], but at least, we hope to get an agreement that this is the definitions and this is the statistics that we're going to go for. And then the data will be the data and that's what we'll be judged on at the end of the day.

And our next question is from Sachin Jain from Bank of America.

Sachin Jain, Bank of America. A few questions, please. Firstly, sorry, just another one on Tresiba CV study. You seem to have a clear view of what you want to do in terms of size, event rates, et cetera. Are there any material topics left open with the FDA post your meeting that still require confirmation or discussion backwards and forwards? Because I'm a little bit confused what's taking you up to 1 year to potentially start in the study. The second question is, we just want a couple -- you commented, Mads, on semaglutide being the most potent agent in Phase III. I think that's what you said. Could you just clarify what your base case expectations for dulaglutide are with the data coming, I think, mid-year. And if that does show superiority versus Victoza, would you consider adding a higher dose for semaglutide in the Phase III study? And then the final question is on the SCALE diabetes, where you've commented that the 3 milligrams was certainly superior on average weight loss than 1.8 milligrams. I wonder if you can just comment whether a 1% difference in average weight loss would be viewed as clinically meaningful? Lars Rebien Sørensen: Thank you very much, Sachin. Mads Krogsgaard, you're on it again here.

Right. Are there any outstanding material issues that you feel we need to have discuss with the agency on designing this trial. I would actually say there are not, Sachin, but we're saying, we will start within the year. There are many things you have to do, including making sure trial supplies are available, that you have all the CMC data available, that you have your clinical feasibility studies done in the different territories with the different investor sites -- investigator sites and so on and so forth. So a lot needs to be done. But I think we've progressed far and we are guiding you that we'll start it within 1 year. We're not saying we'll starting 1 year from now, but within 1 year. Of course, we will act as expediently as we can since this is a highly strategic product for us in the U.S. with both Tresiba, Ryzodeg and IDegLira relying on it, of course. On dulaglutide, okay, very good question. My thinking is that the dula has actually shown surprising -- not surprisingly, both relatively compelling data in Phase III. What we have not received is absolute A1c at body weight effects of the agent in those comparison trials. So I can only refer to what I've seen in Phase II so far. I would expect dula and Victoza to be agents with a similar efficacy profile on glycemia. Whereas, it may well be that dulaglutide, for somewhat short on body weight due to the molecular size of this, you can see, molecular construct. I think you would agree that in Phase II, they didn't see so much weight loss. But that's all speculation and we look forward to seeing the data at this year's ADA Conference in Chicago. And, well, would I consider adding a higher dose in the event that they were to be superior to our agent in their superiority study? We'll take as it comes, but they have only just started there and I look forward to, I guess, next year to see data on that particular program and we'll take it from there. The last one is on the 3 milligrams SCALE diabetes. Well, I would say, Sachin, had I only seen one difference that was specifically significant, and it was a like 1.5-kilogram, or whatever, in favor of 3-milligram dose. What you have to bear in mind here, it's on a host of parameters, it's on also quality-of-life assessments, and we have not even disclosed at this point because they are integrating into a number of different assessments. It's cardiovascular risk markers. It's different kinds of clinical chemistries, and so on, that all turned out in favor of the 3 milligrams dose. So I think in totality, we have a very compelling package here. But it will be much easier to discuss in great detail once we see the 3-milligram data in the SCALE obesity trial and I hope we can discuss that at the half-year conference.

And we'll take our final question from Mark Dainty from Citi.

Just a couple really for Mads. You talked about the matched cohort case study for the pancreatic event, but I may have missed it. Can you just confirm how many patients you've looked at in that study? And then on the same issue, but I thinking about liraglutide 3 milligrams, will you have enough data from the SCALE program to get some comfort on whether there is a relative difference in pancreatic issues between the different doses? And then just the last question on inflammation. I'm just interested, that's a bit longer term, but I'm just interested to know how competitive you think your early stage science is in this area, how strongly you feel it is and whether you feel like there's any need to bolster up in any way? Lars Rebien Sørensen: Thank you very much. This is the last 3 questions. Two of which to Mads. We could also post Mads the last question about how strong our inflammation signs is but I guess, you know the answer to that one. So maybe I'll comment on that. So, Mads, why don't you give it a shot.

Yes. So I'll take the other ones. First, you asked about MACE core study. Now, there are 2 things here. What you do is you establish your cohort of patients taking, for instance -- or taking specifically Victoza who experience a certain condition and then you match them on a number of issues, ranging from socioeconomics to regional distribution in the United States, to gender and behavioral and all kinds of things. And there from a total pool, that differs from cohort to cohort, from the most rare events being cohorts of maybe 10,000 patients to some of the cohorts running up to more than 100,000 patients. And then from those cohorts, you match those that can be matched to your Victoza patients and you will end up -- I would say, typically we'll match cohorts in excess of 10,000 patients -- 10,000 to 20,000 patients matched on each therapy in each group. So even though this is interim, these are actually sizable numbers, Mark. And I didn't allude to that, but I'm happy that you asked the question. When we talk about was it the scale and can we deduce anything -- deduct anything from the scale? Lars Rebien Sørensen: We have enough security and safety observation on 3 milligrams to rule out any excess risk of pancreatitis or even pancreatic cancer, even though we don't believe that there is an association.

I think with pancreatic cancer, hopefully, you would not even see a single because it's a rare event. But let's take that as it comes. I would not expect so. On pancreatitis, you always see some cases. The issue is that you have skewed randomization, typically at least 2:1 on liraglutide, and the numbers of pancreatitis, even though I cannot tell you for sure now, but they will probably be in the teens rather than the 20s as we amalgamate all the programs in the SCALE together. So they are not useful for statistical analysis. You can only do a -- look at the amounts that you have and then put them together with all other insights that you have. And I think the strongest data, hierarchically speaking, are always long-term outcome studies such as the LEADER trials. In clinical terms, the next strongest you can often get is prospectively the size of matched cohort studies and then, obviously, ranking lowest is the low numbers that you get in smallish Phase III trials and retrospectively looking database studies and then, of course, supplemented by all the preclinical and mechanistic data that you have. So I think we actually have a strong and inside pool of knowledge safety base around Victoza and liraglutide, both as it relates to animals and humans, and we will continue to build on that, and obviously, ultimately, we'll have LEADER trial coming out, at the latest in '16, but maybe before. Lars Rebien Sørensen: And so that's all. Lars Rebien here to round off on the comment on our inflammation research and development. We are clearly aware of the fact that this is a very, very competitive area that we're going into. And hence, the innovation part that we demand for our project is rather high so that we have a fighting chance of making an impact in the marketplace. And I would say, from what I have seen of data on the 2 lead projects, Anti-IL-21 and Anti-IL-20, we certainly have some very, very interesting and promising early data. And so, yes, they are innovative and they are competitive, and does that preclude that we should go out and invest further? We made the assessment here that we have a competitive entry here, we have a competitive effort, and we will let the faith of these 2 projects, over the next couple of years, determine whether we should invest more or less in this field. So with that comment, and I hope you [indiscernible]. Then I'd like to thank you, all, ladies and gentlemen, for listening in. If you have any questions, please contact us -- contact our Investor Relations Officers and we are looking forward to communicate with you again as soon as we have some interesting data or the first half-year result, which will be in August.

That will conclude today's conference call. Thank you for your participation, ladies and gentlemen, you may now disconnect.