Novo Nordisk A/S (NOVOB) Earnings Call Transcript & Summary

Hello, and welcome to the Novo Nordisk A/S R&D Investor Call. [Operator Instructions] I'll now hand the floor to our speakers, please begin your meeting.

Thank you. Welcome to the Novo Nordisk Investor Call in connection with ADA. My name is Karsten Knudsen, CFO of Novo Nordisk. And with me today, I have Marcus Schindler, Chief Scientific Officer and EVP of Research and Early Development; as well as Martin Lange, EVP of Global Development. At today's call, there may be forward-looking statements and projections around the future, which by their very nature, clearly, inherently are uncertain and where outcomes may turn out differently. As usual, we anchor our communications around our strategic aspirations 2025, which are covering our strategy execution as a company. Today's focus is purely on the quadrant denoted innovation and therapeutic focus, where we'll be covering our progress within raising the innovation bar for diabetes treatment, developing a leading portfolio of obesity medications as well as building a presence in other serious chronic diseases, especially within cardiovascular disease, NASH and chronic kidney disease. Today, given we do this in conjunction with ADA, we will not be providing specific updates on our Biopharm pipeline. We will do that at other opportunities later on. The agenda for today starts out with diabetes and obesity care, highlights from ADA by Martin Lange, followed by a Q&A session from the audience. From that, we'll move into research and early development with a special focus on cardiovascular disease led by Marcus Schindler and also with presentation by Martin Lange. And also there, we'll be doing a Q&A session following that. So without further ado, I'll be handing the presentation over to Martin Lange.

Thank you very much, Karsten. So first of all, just to start off, I just wanted to share with you our current late-stage pipeline. As you will see across our therapy areas, across diabetes, across obesity, obviously rare blood disorders, but also in other serious chronic diseases, we see progress of our clinical pipeline. Specifically, we are going to start Phase III program across all therapy areas, including in NASH, in CVD and in Alzheimer's disease. These activities have not been impaired, and we do not expect them to be impaired by COVID-19. And therefore, we are progressing our pipeline as planned. This basically means that we expect to have approximately 40,000 active patient years in our clinical development program by the end of '21, moving towards actually 65,000 patients in just 2 years' time. Next slide, please. So I'm going to take you through a little bit of through the force together with Marcus, starting out with what we have on display at the ADA. As you probably noted, we have quite a number of presentations at the ADA. I think it's around 37. And we'll try to give you some of the highlights during the call today. Next slide, please. Starting in the diabetes space, obviously, you've seen our SUSTAIN FORTE trial. It's 2.0 milligram of semaglutide versus 1.0 milligram of semaglutide in approximately 1,000 patients with very high A1c levels between 8 and 10 -- mean baseline being around 8.9 and patients being randomized to either 1.0 or 2.0 milligram of semaglutide for a treatment duration of 40 weeks. Primary endpoint, I was about to say, as usual was hemoglobin A1c, which was our key focus of the trial. Next slide, please. Obviously, gratifying for us to say, a statistically significant and a clinically relevant difference between the 2 doses of semaglutide. In addition to that, we saw a clear impact on body weight and also in terms of the proportion of patients reaching prespecified targets. What was really interesting to us was that -- and obviously, very, very comforting was that increase in dose was associated, as we just discussed, in good increase in efficacy, but with no difference in the gastrointestinal and broader adverse event profile. So as shown on the lower right-hand side of the slide, discontinuations due to AEs -- overall AEs of nausea, diarrhea and vomiting was approximately equal between the 2 doses despite the improved efficacy. Obviously, we take a lot of comfort in that. And as we have also shown at the ADA during post-hoc analysis, dividing patients into subgroups of baseline A1c or baseline BMI, we see the difference between the 2 treatment arms throughout. Next slide, please. Staying in the diabetes space, looking at the PIONEER trial, we've been looking into first PIONEER 1, looking at baseline diabetes duration, so patients having less than 1 year of diabetes in terms of diagnosis versus patients having more than 1 year of diabetes in that diagnosis. We clearly saw that the proportion of patients achieving hemoglobin A1c below 6 was -- and this is maybe not so surprising, better achieved if intervention started earlier in namely patients with very early diabetes. And similarly, in PIONEER 2 comparing glycemic control between oral semaglutide and empagliflozin, we saw that a larger proportion of subjects maintain good glycemic control with semaglutide as compared to empagliflozin. Obviously, very nice data and very nice results for the Rybelsus brand and coming out from both the PIONEER 1 and PIONEER 2 trials. Next slide, please. Obviously, with the very, very strong label that we have with Rybelsus, both the 7 and the 14 milligram showing superior and statistically significant reduction in A1c and a good -- as compared to both placebo, sitagliptin and SGLT-2 inhibitor and good body weight reduction, we also have an ambition to maximize the efficacy that we can achieve with also oral semaglutide. We've already discussed what can be done with subcutaneous semaglutide in the hands of SUSTAIN FORTE. We want to do something similar now in the oral space for Rybelsus. So we are currently in Phase III testing higher doses of semaglutide in an oral formulation, namely 25 and 50 milligram. The intention being that the 50 milligram will correspond to approximately 2 to 2.4 milligrams of semaglutide in terms of both efficacy and safety. So a very nice efficacy and safety profile as already seen from SUSTAIN FORTE with a very attractive safety profile to accompany that. Next slide, please. Staying within the diabetes space obviously, and Marcus will be able to talk about that also in the innovation space, we want to take our diabetes innovation to the next level as well. We have also come to realize, as I think have most players in this space, that monotherapy largely have been maxed out. And therefore, to get to the next level, we need to look at combination therapy. We have or will in very soon future initiate 2 Phase II trials in the diabetes space through that end, first being with the combination of an amylin analog cagrilintide in combination with semaglutide, where we look at the combination versus the monotherapies in a placebo-controlled setting, looking at a fairly short treatment duration. This will allow us to, if successful, to initiate potentially Phase III for diabetes at the same time as we intend to initiate the obesity trials for the Cagrisema molecule. In the combination space, we also have a combination with our GIP offering against a once-weekly molecule looking at the combination of semaglutide and GIP. In this space, we are still looking towards doses and ratio optimization. And we will, in Phase II, investigate a number of different ratios up against the mono components. This is a slightly longer trial than the Cagrisema trial. But again, very high hopes for -- obviously, what these compounds can do, not only in the diabetes space, but also with the potential of very strong weight loss profiles. We had discussed the Cagrisema data in nondiabetes setting, where you've seen weight loss to the tune of 17% over a 20-week treatment period. Next slide, please. Now switching gears a little bit. You've probably, at this ADA, seen a lot of presentations and posters around time-in-range. This is a reasonably new concept, acknowledging that just looking at hemoglobin A1c, just looking at spot blood glucoses or even 7- or 9-point profiles is maybe not enough and not giving us the full picture of good control in diabetes. This is an illustrative slide looking at 3 sort of patients, all with the same level of glycemic control of approximately 8.6 in fasting plasma glucose or 154 milligram per deciliter and 7 in A1c. Now underlying this, if we did continuous glucose monitoring for these patients, we now know that some patients, even with the same A1c or even with the same fasting plasma glucose, have very different time-in-range profiles. Time-in-range basically depicting how much time does the patient spend in what we would call good glycemic control. I've recently heard an investigator saying it's sort of like the speed limit. If you are to stay within, let's say, 50 and 100 miles per hour, that wouldn't do in U.S., but it works in kilometers per hour, then that's your goal. If you go too higher or too low, you have problems. In this example, we clearly have 1 patient spending way too much time in high glucose levels leading to increased risk of late-stage complications, but also way too much time in low blood glucose levels leading to risk of hypoglycemia basically. We have an example that sort of in between. And then we have 1 patient being in exactly the range where we want the patient to be. But without the technology of continuous glucose monitoring, without the digital tools and the technology to help us interpret this, we wouldn't have those insights. Having those insights, it allows us to better target and tailormake treatments for patients. It allows us to better understand the disease. And therefore, we see this technology advancement as a major advancement also in the treatment of diabetes. Next slide, please. Specifically in the insulin space, obviously, introducing new insulins, we want to make sure that we do see an upside, not only on A1c, not only on fasting plasma glucose, but also specifically on time-in-range because this is where we know that we can also start predicting outcomes for our patients. So as you know, we've conducted Phase II for insulin icodec looking at different titration algorithms and comparing to insulin glargine in a 1:1:1:1 randomization in Phase II and a 16-week study duration. In that space and looking at titration B, we see actually that being on icodec is associated not only with the statistically significant but also clinically relevant 10% more time-in-range. And you can probably imagine throughout the day being 10% more time-in-range. It's very attractive to the patients, obviously, in everyday life, but also in terms of the outcomes. This gives us very high hopes and very high aspirations on behalf of the icodec molecule. Next slide. And obviously, we intend to investigate that in Phase III, and I'm going to show that in just a minute. Specifically, also for this ADA, we've been able to look more into the risk of hypoglycemia with a once-weekly molecule. This is, in fact, a recurring story with the introduction of the longer-acting molecules, for example, Tresiba. There were questions around if these work longer is the risk of prolonged hypoglycemia, that's also present. We could show a Tresiba that, that was not the case. Actually, duration of hypoglycemia with Tresiba was comparable to that single, for example, both Levemir and glargine. But based on what we've seen also for icodec, which is now going from once daily to once weekly, we are super, super happy to see that the risk of having prolonged hypoglycemia is not there. So basically, the timing of a hypoglycemic event is the same, both obviously in the different icodec titration arms but, more importantly, comparable to that of insulin glargine. And that goes for both overall but also nocturnal. Very, very comforting for us to see. Another way of looking at it is obviously -- and this is based on the CGM data, time spent below 54 milligrams per deciliter which is, in fact, down right hypoglycemia. And again, here, we see no difference between icodec and glargine, suggesting that even in the occasions where we do have hypoglycemia, the risk of staying too long in that hypoglycemia is not there. Again, very, very good data for insulin icodec and, obviously, adding to our aspirations for that molecule. Next slide, please. To me, very, very exciting, we are in the progress. We've actually finalized the recruitment for ONWARDS 1, which is our pivotal and obviously also longest trial. We are well on track for ONWARDS 2 to 6 and actually none of them being behind. Actually all on time in terms of recruitment, actually making us really reasonably adamant that we will be able to finalize and report on these trials in a timely fashion. So we will be able to do regulatory submission either late '22 or early '23. For all of these trials, we obviously apply a treat to target, apart from ONWARDS 5, which is a real-world evidence trial, where we're basically using digital tools, allow the patients to control their own treatment together with digital support, basically a connected device, a connected BGM and then obviously an app to guide the patients. We have employed CGM in ONWARDS 1, 2, 4 and 6 to support our assessment of also time-in-range. Obviously, we are very, very excited about insulin icodec and we'll be clearly excited when these data start to read out in '22. Next slide, please. Now moving into the obesity space for a brief while. We talked about a couple of weeks ago in connection with the very nice approval of semaglutide 2.4 milligram or Wegovy between friends in the U.S. We shared with you the STEP 5 data. STEP 5 is basically a trial dedicated to look at the sustainability of weight loss with semaglutide. As you know, with the current treatments, we do tend to see, over time, after approximately 1 year, a waning off of the weight-lowering effects of the treatment. And we set out to investigate if that would also hold true for semaglutide. Very comforting for us, almost boring for us to see that even after 2 years of treatment with semaglutide, we still see the same 17% weight loss that we have reported for 68 weeks of treatment. Very, very consistent across all the studies that we reported on semaglutide 2.4 milligram so far. Obviously, comparing to placebo, this is not only clinically relevant but also highly statistically significant. As almost pedestrian or boring, the same 40% of patients achieved at least a 20% weight loss even after 2 years of treatment. And obviously, we are very, very happy and this is also becoming a feature of semaglutide, an attractive safety and tolerability profile. Exciting for, obviously, Marcus and myself to see obviously that we see improvements in lipid profile, but also, importantly, of C-reactive protein as a marker of anti-inflammation effects of semaglutide. We believe this to be important for semaglutide in the NASH space, in the cardiovascular space, but also in the Alzheimer's space as part of the mode of action for semaglutide in these spaces. And also these findings have been consistent across all the STEP trials. Next slide, please. In STEP 8, which is a trial that has not yet been reported, so this is basically the first. We set out to compare liraglutide to semaglutide 2.4 milligram in a placebo-controlled setting. So both treatment arms were placebo controlled. For the sake of simplicity, we put placebo in this slide, but both treatment arms were placebo controlled. In this space, again, we see a 17% weight loss with semaglutide and we expected approximately 7% weight loss with liraglutide 3.0 milligram, basically talking to the expected effects of the 2 molecules. What is really, really nice, obviously, and I don't have time to show that here, is we also see effects on glycemic control even in this nondiabetic population. And we also see the impact on lipid profile and inflammatory markers as we just discussed for STEP 5. So a very attractive trial, also considering that semaglutide comes out with this very attractive offering, including a benign safety profile and given that semaglutide will be priced to the level of liraglutide then with a more attractive clinical offering. Next slide please. Now in STEP 1, and this is obviously a study that has already been reported, we've done a sub-analysis showing the impact of glycemic control in a nondiabetes population. As you see in the right-hand side of STEP 1, and you probably remember STEP 1 was our pivotal obesity trial. Approximately 2,000 patients being randomized to either placebo or semaglutide. And in this space, we see that pretreatment, approximately 45% of patients in the semaglutide arm and 40% in the placebo arm had prediabetes. After 68 weeks of treatment, only around 9% of patients in the semaglutide arm still have prediabetes. And that is to be compared with 26% having prediabetes and 1.6% having overt type 2 diabetes. Obviously, very, very interesting data in and of itself. What we've not reported so far, but what I can display here is that we actually did an extension of STEP 1, where we observed the patients off treatment for another year, so 52 weeks up until 120 weeks. And in this space, we actually see that the legacy effect of having been on semaglutide is maintained. So more patients stay in the prediabetes -- sorry, in the nondiabetes states with semaglutide than what we see for placebo. So not reverting to sort of a diabetes risk baseline despite the fact that they actually start to regain weight. So very, very attractive and a good outlook also for semaglutide, in particular when we see -- start to see similar data coming out of the SELECT trial. Next slide, please. In STEP 2, we looked at in improved glycemic control and other markers in a post-hoc analysis post the 2 -- knows the 2 was our weight loss trials in diabetics. What we saw was a marked improved in waist circumference. We saw improvement in lipid profiles. We saw, again, improvement in inflammation. And importantly, we also saw improvement in health-related quality of life questionnaires, as measured by both SF-36 and the IWQOL. Next slide, please. Overall, we were very, very happy with the efficacy profiles that we've seen coming out of STEP 1 through 5 and STEP 8. We have seen a very, very robust, but also consistent weight loss. We have seen robust and consistent improvements on cardiovascular biomarkers, including inflammation biomarkers. And we've seen this with a very attractive safety profile that we believe is starting to become a hallmark for semaglutide. Just very briefly here. I hope this is not my timer. We're just showing here the discontinuation rates for semaglutide in STEP 1 and 2 as well as for placebo. And as you will see, first of all, these discontinuation rates overall due to adverse events, but specifically also due to GI-related adverse events, are very low and reasonably close to the discontinuation rates observed for placebo. This discontinuation of treatment has not led to very high withdrawal rates. Generally, we see withdrawal rate from our trials in the tune of less than around 5%, which is in an obesity setting, a testament also to the acceptance of the drug. Next slide, please. So overall, we believe that we have had an interesting presence at the ADA with a number of different abstracts and presentations. We continue to raise the bar in the diabetes space with more benefits of semaglutide, as specifically demonstrated by SUSTAIN FORTE. We continue our insulin innovation, and we are very, very excited both with the current data from Phase II, but also with the outlook for icodec in Phase III. We're going to initiate 2 Phase II trials to further raise the innovation bar in the diabetes space, one with the combination of cagrilintide and semaglutide and the other of semaglutide in combination with GIP. And finally, we show in the obesity space a profound weight loss and -- with the magnitude and duration that will obviously be explored further, also the impact of which in the ongoing additional STEP trials, but also in the SELECT trial.

Thank you, Martin. That takes us to our first round of Q&A, operator. [Operator Instructions] So operator, please move to Q&A.

Our first question in the queue comes from the line of Wimal Kapadia of Bernstein.

Wimal Kapadia from Bernstein. Just got only -- we have 1 only. Can I just ask about how you think about the fixed dose combination approach for GLP-1 and GIP that you're developing versus the dual agonist approach? You mentioned the balance, but does the fixed dose really drive better safety outcomes? Or is there potential for superior efficacy if you get the ratios finally balanced? Just curious what you see as a differentiator given you're several years behind your main competitor.

Thanks, Wimal, for that question. And this goes to you, Martin. So fixed dose versus dual agonist, what's our considerations there?

So it's a super good question and obviously also an approach that -- or a question that we asked ourselves. We have tested this both in the preclinical space, but also in the clinical space where we have evaluated a number of different ratios. And clearly, we believe that specifically for the combination of cagrilintide and semaglutide, but also soon for the semaglutide GIP that we've sort of hit a sweet spot where we've optimized efficacy, but we've also managed to do that without compromising on safety. So specifically on the Cagrisema combination, you saw this rather dramatic increase in efficacy, 17% weight loss in 20 week, which is to be compared with the 17% weight loss we see with semaglutide in monotherapy in 68 weeks. But this comes with no increase in tolerability and no increase in adverse events. So we basically get the improved efficacy without compromising at all on safety. Obviously, we had to prove this in Phase III as well. But we do believe that with our approach, we managed to find a ratio that optimizes the best of 2 world.

And that comes from the line of Simon Baker in Redburn.

Just on that point about the combination, I just wonder, and I realize it's very early, but what data you can share so far, even if it's preclinical -- if you have any preclinical data against tirzepatide, that would be very helpful.

Martin, what's -- what do we know at this point versus tirzepatide on our combination?

So in the clinical space, and I believe in the preclinical space, we don't have head-to-head evaluations, neither we know nor Lilly has that at this point in time. So at this point, we had to rely on modeling. And I mean, we have decided not to fully share our modeling of specifically the Cagrisema combination. However, we do believe even with our most conservative modeling that we will, at the very least, be able to stack off, if not be superior, to what is currently out there.

That's from the line of Michael Leuchten of UBS.

I was just wondering on timing on your cagrilintide Phase II. Would there have been a possibility to make this straight into a Phase III? And the reason I'm asking is Lilly seems to be talking about moving the triple agonist into Phase II this year, which, then, if successful, may then mean they're not that far behind your combination. So just why not be more aggressive with the timing I guess is my question.

Thank you, Michael. That's a very good question. the CFO thinks. So Martin, can we move faster than what we have seen here?

It's a super good question. And obviously, we're taking a lot of comfort in that we've spent the last couple of years not only in decreasing our white space substantially, but also taking smarter in terms of how we go about development. That being said, we always have to do an evaluation of the data and the knowledge that we have and what needs to be done in order to increase our Nordisk space before we go into Phase III. We have good examples, [ Salsify ] being one, icosema being another, where we go directly into Phase III from Phase I. However, with the Cagrisema, we have assessed that we need a little more data to fully be able to design Phase III. That is not going to impair our time line substantially because we have shown that we can both recruit and also conduct our trials very, very fast.

And that's from the line of Michael Novod at Nordea.

Just a question to STEP 2 and also your STEP symposium or the STEP symposium those to the other day at ADA where there was a discussion between the panelists around use of higher dose much faster. So how do you see the use of sema 2.4 potentially also in diabetes because it was clear that Rosenstock at least said it was time to move on to much higher doses in diabetes based on the data that's been presented for this dose in diabetics. So how do you see that sort of interact with Wegovy in obesity versus the use of this high dose in diabetes? Of course, also bear in mind, you have 2.0 going to the market soon.

Thank you, Michael. So Martin, this is something we debated when we designed the trial. So 2.4 obesity dose versus 2.0 dose in diabetes for semaglutide.

Yes. So first of all, I mean as with everything else, this is dependent on an assessment of the individual patient. If the key focus is weight loss, I would go with Wegovy 2.4 milligram. If the key focus is diabetes control, I would go with Ozempic, either 1 or 2 milligram. And again, I think, clinically speaking, I don't want to contradict Julio Rosenstock. Obviously, again, it has to be based on the individual assessment. There will be patients who are on optimized glycemic control at 1 milligram and with a very attractive weight loss. But there will also be patients who either need a little bit more or need longer sustainability. So the option of adding a higher dose, which makes the -- both the 1 but also the 2 milligram of Ozempic a very attractive offering. But I think it's super important to distinguish between the diabetes focus or the glycemic focus and the weight loss focus. Obviously, the 2.0 milligram does also provide a very nice weight loss. But if the key focus is glycemic control, 2.0 milligram is the way to go.

Thank you, Michael. Thank you, Martin. And as you know then we're also pursuing high doses in our oral semaglutide offering as well with the 25 and 50 milligrams, so we are clearly looking in that direction.

That's from the line of Peter Verdult of Citi.

Peter Verdult from Citi. Martin, just I guess my question, the sema-cagri dual chamber devices, are you on track to enter Phase III with the dual chamber device? Yes or no? And then only if you want to answer it, just when will we get data from some of the other obesity analogs that you're prosecuting, PYY and so forth? Is there anything coming on the horizon?

So first answer is yes. And second answer is these studies are currently ongoing. And when we've seen the readout and assessed them, you'll be almost the first to know.

Thanks, Pete. Very clear and short answers from Martin so -- and close questions. So thank you for that, Pete.

That's from the line of Emmanuel Papadakis of Deutsche Bank.

Maybe one for you, actually Karsten. If you could just give us a quick update on the latest in terms of COGS at least on a relative basis for 50 microgram doses? So where do you think the gross margin for that is likely to come in relative to the rest of the franchise given some presumable process improvements? And then just a quick clarification. If I heard earlier, Martin, I think you said you thought 50 mg would be clinically equivalent to the 2 to 2.4 milligram doses subcutaneous semaglutide. Did I hear that correctly?

Yes. Emmanuel, thank you for that question. And that's clearly something that we, of course, evaluate before we initiate the -- we did evaluate before we initiated the high-dose oral semaglutide trial. And what we've been saying previously has been that we expect for Rybelsus that we will get to group average margin by our strategic aspirations period for -- in terms of gross margin for Rybelsus. So -- and what we're doing at the same time is, as we've been talking to before, we continue to invest in our oral platform in terms of oral formulation and our SNAC platform. So that is a Novo classic that, as you've seen with our insulins and the performance for many years. We do that with our oral platform also. So clearly, without guiding specifically on the gross margin for the 25 and 50 mg, then we would not have initiated high-dose trials either in diabetes or the obesity oral trial without being able to get to a place with an attractive gross margin for those products. And Martin, then the second question for you.

Yes, I think you're exactly right. So in the diabetes space, we expect the 50 milligram to correspond to the 2 milligram subcutaneously. And in the obesity space, and again we see these as 2 different disease entities. We expect to see the 50 milligram correspond -- approximately correspond to the 2.4 milligram in terms of weight loss. So in that space, acknowledging that the oral formulation has a slightly higher variability, our model suggests that 50 milligram in diabetes will correspond to 2 milligram and 50 milligram in obesity will correspond to 2.4 milligram.

Thank you, Martin. Thank you, Emmanuel. And then it's time for the last question before we move on to Marcus and research and early development. So last question, please.

And that's from the line of Richard Vosser at JPMorgan.

Just on the icodec data, it looks as though you titrated slightly more aggressively than the Lilly weekly in their Phase II trials and potentially got a few more patients spending or more time of patients in range. Could you maybe contrast the benefit? Is that right? And is the benefits of the product to icodec versus B as you see it at this point?

So I would be super concerned to compare another investigational drug. What we've seen so far makes us very confident that icodec has a clear edge, so both in terms of time-in-range, but also in the maintenance period, we actually saw a lower risk of hypoglycemia with icodec than what we saw with the comparator once-daily drugs, specifically glargine. That being said, obviously, with the titration that we had, we saw in the titration period that we need to do some work. And we basically spent the combination of our Phase II data, as you also saw we had 1 study with a loading dose when switching from another insulin and so on. And the combination of all the data that we had allowed us to do what we think is going to be the good but also ambitious titration algorithm, balancing good efficacy, but also good safety in terms of risk of hypoglycemia.

Thank you, Richard. Thank you, Martin. This ends the Q&A session and it takes us to the session on research and early development. So over to you, Marcus.

Thanks very much, Karsten. Go over to next slide, please. My name is Marcus Schindler. Obviously, as we haven't met either in person or probably at any time in my new role over the last couple of months, I just wanted to allow myself a little bit to introduce you to research and early development, because we are in the midst of transforming the way we do drug discovery on a number of avenues. And I think there's a lot of exciting pieces here I actually want to share with you rather than spending too much time on individual assets, but we will be talking about some exciting news on our cardiovascular asset because that is sort of the next frontier, obviously, that we're tackling. First of all, maybe to the left on your slides, we're moving really from a world where we have based hypotheses largely on observations based on preclinical animal studies to something what we call human-centric drug discovery. So with the power of computing today, human genetics really, I think, is moving into the center piece, and the initial disappointments on some of the GWAS data actually are getting resolved with the granularity, we can interrogate some of those data. So that's point number one. Second is we have now large databases from clinical trials, biopsy samples, soluble biomarkers that can be interrogated and all of that to build actually new hypotheses for novel treatment paradigms. And last but not least, to use one of the buzzwords, obviously, big data, interrogating real-world evidence data will also actually help us to come up with novel ideas. And all of this, if you then lump it together with human-centric model systems, where we're using organoid systems, for example, to actually test our drugs, leads us to a world where human data and human model systems are at the very core of everything that we do. And I hope you agree that it's actually quite distinct from where we were. And I would fundamentally argue that's a better place to be because we're taking out early attrition very early on in the pipeline rather than pushing observations later on into clinical development to then deal with the outcomes. With several thousand employees working hard really on innovation, and it's not something to boast about, but simply also to remind all of us on our leadership position. And I've worked in a number of pharmaco companies, but clearly, in Novo Nordisk, we are dedicated to the disease areas and it's a smaller number than some of our competitors. And I think that also gives us a very deep insight into the science and a long-term outlook into the biologies we're working with. We're moving from an organization that has been largely centered around Denmark and being located into Denmark across the globe. And why is that a good thing? Because we're tapping into the talent pool that is globally available in other epicenters, obviously, of innovations such as Boston. We're present now in the Bay Area. We have a site in Seattle, very active in Oxford and in Beijing. So I think that gives us a really fundamentally good base of cutting-edge scientists and people working on various technologies to do the work that we need to do. We're also moving beyond, I would say, what is maybe a more traditional, functionally structured way of working into one where we have a number of ways of working in operational units, so-called transformational research units. You probably have heard about our stem cell unit, with a ring-fenced budget and a level of autonomy within our framework that actually allows them to move incredibly fast, test hypotheses early on. But at the same time, also take accountability and be sort of mini CEOs within our company. Maybe one thing you haven't heard so much in the past is a network of partnerships, because largely all the good things that Martin have showed you were and are the product of in-house innovation, but we firmly believe that the next wave of innovation will actually come to dedicated specific work with partners. Those could be academic institutions, could be biotech and actually peer pharma companies. And we believe that is one way and maybe the best way to tackle, in particular, number of unknowns that we're dealing with. Now we have a large number of projects in the pipeline. Now 90-plus doesn't actually tell you a lot. And I have to say what actually energizes me more is that a significant portion of those projects deal with first-in-class novel mode of action. And I think that really puts us in a position that we have a significant number of proprietary drug targets where we feel we can be in a leading position. The second piece is we see a larger diversification of the molecular entities that we're using. Of course, you know us as a company leading in the biologics space focused on peptides and proteins. But we're going way beyond this. We've touched on the oral biologics leadership, and that is continuing, I can assure you that, full steam ahead. But we're also entering into the field of siRNA therapeutics to address intracellular targets and silence them more or less completely. We're using stem cells to move us into the regenerative space. And we're actually generally broadening out to other tools, but not leaving what we see as our fundamental scientific core source. There are also currently modalities that we're not working on. And all of this obviously comes together that we're not only staying and continuing to innovate in diabetes and obesity, but we're moving beyond. 2/3 or more of people living with diabetes suffer from cardiovascular disease. A significant portion of them will see cardiorenal disease. So branching out into cardiovascular specific and also chronic kidney disease, I think, it's actually a very natural journey. Through early work also in the stem cell unit, we are moving towards sort of the regenerative spaces, which take us a little bit outside of our therapy areas. We're entering into Parkinson's disease. But we obviously also have semaglutide now in Alzheimer's disease, which I think is a tremendous opportunity to pursue. And we're not speaking today because this is the ADA call about the exciting pieces obviously in the rare blood and rare endocrine disorders, which I think gives us actually a very, very nice angle to start experimenting with and engaging in gene therapy and gene editing. Next slide, please. So many of the things actually I talked about have started, I think, with GLP-1, semaglutide in particular, which has really paved the way with seemingly a plethora of biological pathways that we're addressing. And for those of you who are old enough to witness when these molecules were started, and I saw today, obviously, [ Daniel ] holds the Banting Award for Incretin Biology I had started decades ago, but if you imagine today, what we know about GLP-1 and where we see the fundamental effects we're seeing in a number of organ systems, it has really opened our eyes to the opportunities. On the right-hand side, however, giving you some evidence that we're not only broadening out in disease areas, but we're actually doing this ASU partnership and we're using specific platform collaborations to actually broaden our ability then to address the targets that we want to pursue. Sort of leading the pack maybe here, the Dicerna collaboration, which we signed a strategic cooperation with Dicerna on siRNA technologies, progressing very, very well, very productive collaboration. We're taking first steps in gene editing with colleagues at bluebird bio in the hemophilia space most recently, and I'll come to that in a minute. We signed a deal with a Japanese biotech company Heartseed. And maybe the most prominent some 3 years ago around this time was our acquisition of Ziylo and the legacy company Carbometrics, which gives us access to the glucose sensor part of one of our tracks on glucose-sensitive insulin. And of course, you're all aware of the acquisition of Emisphere, which really paves the way now to -- to blast sort of oral biologics and generate better and better technologies here. We'll talk a little bit about cardiovascular today, in particular, on ziltivekimab coming from the company Corvidia. But then as one fantastic example, I think, of a peer pharma-to-pharma collaboration, I just wanted to highlight the Gilead collaboration and how together we are tackling NASH. Next slide, please. And of course, in cardiovascular disease, there remains a significant unmet medical need. Despite standard of care, despite actually all of us in the field making good advances to what we're also seeing from the SGLT-2s, but also from our own molecules, unmet medical need remains. And next slide, please. Heartseed is an interesting collaboration for us because it basically took -- or ticks 2 boxes for us. One was to have a presence in the cardiovascular space. On the other hand, maybe even more prominent, to collaborate on stem cell-based therapies. So Heartseed over probably 2 decades of, no pun intended, hard work has actually come up with a rather unique iPSC preparation for cardiomyocytes. And they've really optimized the protocol to come up with cardiomyocytes, which from all we have seen look very benign. They show no abnormalities in -- once they're implanted into the heart. And they've actually come up with a way to implant those cells and they form little spheres, which I don't want to say guarantees, but has probably a better chance to survive in the heart and stay there for a long time than we have seen with previous versions of stem cells. And we're actually very excited that we can move, hopefully, into a first time in-human trial very soon, and are pretty much on track to do this in the near future. Next slide, please. So moving from a field where stem cells will really help patients to not having to go on a transplant or at least prolong their time to a much more fundamental biology, which is inflammation, right? And maybe it sits in line here with the triglycerides and the cholesterol. Maybe we should have actually a bar that cuts across because we do believe it is a fundamental principle that is ongoing. What we're going to zoom in here now in a very particular patient population, next slide, please, which will be addressed by an IL-6 block. And this is, you remember, I spoke briefly about this last year at our call when we just acquired Corvidia. This blocks IL-6, the ligand, not the receptor, we believe that actually gives us a better profile, a better therapeutic window and not to be concerned so much with side effect. And the news you have seen probably at ACC is that ziltivekimab does exactly what we were hoping for with very high potency. It basically blocks the hsCRP to a large degree. And maybe as a fun fact thrown in, before we saw the data, we had made an internal checklist to see what would we expect, what would look great and what would look outstanding. And I can share with you, this was the outstanding box, because all 3 doses actually delivered significant reduction. And what we're not showing here, but you've probably seen it in the publication, the reduction is sustained. And that, of course, is very, very important. We've also seen a very benign side effect profile. Now these are small -- relatively small patient numbers, it's a Phase II trial, but it basically ticked all the boxes we could expect to see from a Phase II trial. Next slide, please. And that actually gives us confidence to move forward with this drug, and I'll let Martin speak to the ZEUS trial.

Yes, very briefly. So as Marcus rightly said, we've seen some super exciting and interesting Phase II data, both from an efficacy but also from a safety perspective. It has allowed us to identify what we believe to be the optimal dose. And we will actually progress that directly into our cardiovascular outcomes trial. So the ZEUS trial is going to start recruitment in Q3 of this year. It's going to be 6,200 patients being randomized to either ziltivekimab 15 milligram on top of standard of care versus placebo. This is a fairly advanced cardiovascular population as compared to what we have seen previously in our trials. So all of these patients will, I was about to say, as usual, have established cardiovascular disease. But on top of that, they will have chronic kidney disease Stages III and IV as well as having an underlying level of inflammation, which leads to a reasonably high risk of cardiovascular events. This is also why we've allowed ourselves to have a smaller sample size than what you've seen previously in our hands. Typically, we are around 10,000 patients in our outcome trials. In this specific space, because of the expected high incidence rate of MACE, we can make it a little bit smaller. We still had to have exposure for a certain period of time to live up to regulatory requirements, but we do expect a fairly fast recruitment and a reasonably fast conduct of the trial. Back to you, Marcus.

And obviously, the job isn't done just by dealing with inflammation. I think it will help us, but there's still unmet medical need, in particular on high cholesterol. Next slide, please. In the interest of time, I don't think we need to walk you through how PCSK9 works. You're very familiar with it. What you also know is obviously that antibodies can block the PCSK9 function. And you are, of course, aware that antisense or siRNA can actually block PCSK9. What you might have not known so far is that actually our really genius technologist chemist designed a peptide inhibitor of PCSK9. And that is a rather unique molecule, and we'll share some of the details in the peer-reviewed application when it's ready. Now we obviously wanted to test, would a peptide inhibitor against PCSK9, something that nobody to our knowledge has ever done before, actually work? And we designed a small Phase I trial here. And then the next slide, please. I can share with you that we actually got very nice data. We saw a dose escalation here between 150 and 250 milligrams of this inhibitor. And the very interesting thing was not only that we actually could reduce LDL-C significantly both in a healthy cohort, but also in patients with elevated LDL levels significantly. But the interesting part is that this reduction was actually very, very long lasting. And it happened at a dose that would actually make us comfortable that we could turn this peptide into an oral formulation. And that we feel is the rather unique value proposition going into this market, that we're moving beyond an injectable space and actually provide an oral offering to the patients out there, with a similar, hopefully, efficacy and side effect profile as we have currently seen. And we obviously are very keen to test this profile in later-stage clinical trial. Martin?

Yes, it is and here, obviously, we are trying to be smart because obviously going directly from subcutaneous to all could present here warrantees in other Phase I trials. We do believe, however, we understand the both PK, but also properties of going through a very, very nice. And therefore, based on modeling here, it may -- sorry, I had my microphone off. First of all, I'd like to say it's super nice to get the oral aspect out in the open. I've been talking about our PCSK9 offering for some time now. And some of you have asked what do we have to offer as compared to anyone else? From our perspective, we obviously believe that an oral offering will be very, very attractive in this space. Also going back to the question I received before, we are actually trying to be a little bit smart here because it would be natural to say going from subcutaneous to oral, we would have to repeat Phase I in the oral space. We do, however, believe that we have a good understanding of the PK and PD properties of our molecule. And therefore, we've modeled the expected therapeutic doses and we take them directly into Phase II in a combination -- sorry, a comparator trial where the active comparator is going to be repaired. Based on that, we intend to be very fast into Phase III, also signifying the fact that we have very high aspirations for this compound.

And maybe just to really briefly wrap up. I hope you received some interesting data that to showcase and just to give you a few examples, we are venturing into the cardiovascular space. Some might argue we're already in cardiovascular because our drugs show fundamental impact on cardiovascular disease and death. So we're just moving into this space in a broader way with a larger offering into our pipeline. Zilti, I think the first one out of the block with a Phase III trial, but we're obviously also very curious to see how the oral PCSK9 will evolve. And many of those things, we didn't talk about statin, the ApoCIII compound, which is in pipeline, Heartseed. So you will see many more of our programs actually coming through a partnership model to push into this disease area for us. And just to thanks for your attention for now.

Thank you, Marcus. Thank you, Martin. Clearly, we see a lot of activity not only in development but also in research and early development. So it was a pleasure sharing that with the audience today. So just very briefly to summarize. We believe that we are clearly raising the innovation bar in diabetes. We've been discussing icodec, Cagrisema and the sema combination with the GIP. You see our pipeline in obesity, both with the semaglutide 2.4 as well as Cagrisema. Biopharm, we continue to progress, and we'll cover another day. And then finally, as Marcus and Martin just covered, we have a lot of activities in other serious chronic diseases. Here, we covered cardiovascular disease today. So very good progress on the innovation and therapeutic focus against our strategic aspirations 2025. So with this, we move into the final Q&A, 15 minutes, and we'll proceed with one question per person. So operator, we're ready to take the first question.

And that's from the line of Mark Purcell at Morgan Stanley.

One for Martin. Martin, in terms of glucagon receptors, I wonder where you are now in terms of that target. It's a target you previously looked at, but turned away. It seems if you get the right ratio, like Altimmune have now in the NASH trial. It appears to be a very attractive target. So just wondering where you are with that. And it's just a follow-up to about 3 questions earlier on, on GIP. You were doing a Phase I trial with your GIP plus sema. And you're now going to Phase II with ratios of between 1:1 and 1:9. Obviously, that's a pretty broad range. So just trying to understand what you learned from the Phase I trials.

Great. Thank you, Mark. So Martin, the first one on our view on addressing glucagon receptors. And then the last one on what we learned from Phase I since we're starting Phase II now on GLP-1, GIP.

Yes. So thank you very much for those 2. On the glucagon, obviously, as you also mentioned, we both had our own version -- our previous version of coagonist. We also looked at a triagonist at one point. That was in a once-daily setting, whereas the other one was in a once-weekly setting. As with everything else, this is a balance between efficacy and safety. And in this sort of dual-triple agonist context, finding the right balance can be difficult. And at least in our hands, we actually saw a reasonably good efficacy also when combining with the glucagon. However, it did not stack up to what we saw with Cagrisema. And given that, as we already discussed, we saw a very nice safety profile with Cagrisema from a risk-benefit perspective. It really didn't fit our pipeline. We do believe that we can get -- optimize the efficacy with Cagrisema and at the same time have a substantially more attractive safety profile.

And then GLP-1, GIP.

Yes. Let's just say that we saw data that would suggest that it would be reasonable from an efficacy, but also from a safety perspective to go into Phase II.

That comes from the line of Peter Sehested of Handelsbanken.

Relating to your comment about increasing number of patients from 40,000 to 65,000, that's a 50% increase. Of course, you get the obvious question despite targets within your strategic aspirations period. Should we anticipate a significant increase in R&D costs going forward? If not, could you please explain the sort of efficacy or efficiency processes that is driving your ability to maintain cost despite a significant increase in patients and whether it has some read across to say some of the new therapeutic areas that we're entering into in terms of R&D efficiency? And if time permits, it's not just an optional curiosity question of whether AlphaFold is a game changer and whether it's something that you can benefit from?

So Peter, thank you for those questions. So if we start out with the impact of the number of active patients on clinical trials, then clearly, that is driving up costs. So yes, we are going to spend significantly more on R&D in the years to come. We are in the fortunate setting that we have a top line, which is showing very, very solid momentum these days also and this year. So as a company, I think it's clearly the right thing to do to continue to invest in R&D. And as I was out communicating already at our latest Capital Markets Day in late '19, you should expect our R&D ratio to be increasing gradually over time. So you should not be surprised if our R&D investments are growing at a faster pace than sales. Of course, we're looking for -- and on your second part of the question, on efficiencies, we're looking for efficiencies across our value chain. So of course, part of the equation of a broadly stable operating profit margin is, of course, efficiencies across the value chain. And we get a very nice gearing also in R&D on efficiencies, both in research and development from, I'd say, automation and digitalization being the, I'd say, the most prominent enabler of continued R&D productivity. But perhaps, Martin, if you give a few words on R&D productivity.

Absolutely. Obviously, Marcus should speak to research, but I think it's a fair statement that Marcus has an ambition of approximately decreasing per asset research growth by 50%. So super ambitious. And we've obviously tried to look for the same. So both in terms of the resources that we spend, but also the money that we spend, we've been looking at efficiency gains over the last 3 years. We'll continue to do so. Just to give you the example. In 2018, I had 1,200 employees to conduct trials to the tune of 12,000 or 15,000 patients I think we were in 2018. Today, I still have the same 1,200 people conducting our clinical trials. And we are, as we just discussed, at 40,000. So a substantial efficiency gain already there. To Karsten's point, this is about digitalization. But it's also around thing about which countries, how to do procurement and so on. So some would say pedestrian, but obviously, in our heads, we, at this point in time, could do approximately twice as much today as we did 3 years ago for the same amount of money. We'll continue that journey. To your other point, obviously, moving into new disease areas, we have to look at the cost. And typically, we talk about cost per patient. I can tell you already now, we had a wide range cost per diabetes patient is substantially different than cost, for example, for a hemophilia patient. And that also goes for our new disease areas. So for example, a cardiovascular patient is very much, so to speak, cheaper or less costly than a patient in the Alzheimer's space, which is comparable to diabetes obesity. But then we see NASH patients being substantially more expensive. So again, across what we do, we had to differentiate a little bit more, but we also had to accrue the efficiencies that we see. I do want to do a shout out to our model of not using CROs because that in and of itself is cost-efficient. We have our own people doing our clinical trials that allows us to plan better, but also gives us better efficiencies. And it also gives us better quality from a regulatory perspective. So we are very, very happy with that model. On the last question, I think I will leave that to Marcus.

And I think that was around AlphaFold. So basically, the machine learning to predict protein structure and protein folding. And of course, we see that as a really important scientific advancement, and I think the field is very excited about this. And in a way, it ties very nicely in with efficiencies because, obviously, we're very interested in how we can use AI in drug discovery and drug design. And I think that will play a very crucial role. We see early biotechs now coming up, which potentially disrupt basically this field. And that is something we're not only aware of but where we're actually quite active to see where we make our position.

That's from the line of Simon Mather of BNP Paribas.

Just quickly on the oral PCSK9 peptide using the SNAC technology, is there anything we need to be aware of, again, with respect to taking on an empty stomach? Just when is the same as Rybelsus. And just a clarification on Mark's question. Am I right in understanding that you're not going to go down the triple combination therapy, including glucagon? And just on your comments that Cagrisema is super clean, just wondering why that was the case.

All right. So dosing conditions of oral PCSK9, Martin, triple agonist. And what was the last one?

It was on Cagrisema.

I'm sorry, why is Cagrisema clean?

Yes. So we do expect to see dosing conditions similar to what we've seen with semaglutide. Obviously, that is part of what we will investigate in our Phase II trial. So maybe a little too premature to speculate. It is correct that we do not in our current pipeline nor in our portfolio have a triple agonist and we have no plans to do so at this point. And finally, for Cagrisema, semaglutide and amylin are acting on 2 different parts of the brain. And what we have seen so far is that amylin or cagrilintide in monotherapy was associated with a very benign side effect profile. And that stayed true when combining with semaglutide. So we basically saw a side effect profile that was similar to that of semaglutide 2.4 in monotherapy. So these patients had side effects, but no more than what we saw for semaglutide 2.4 into monotherapy.

And that last question comes from the line of Carsten Lønborg of SEB.

I just had a question on ziltivekimab because as you say, target inflammation and you say it's a driver of increased risk of CV disease. But how certain are you of this actually? Because we, of course, have seen a lot of indications from cancers, et cetera, that this is the case. But could you try to sort of describe a little bit more in detail on how certain you are that this is actually the case, but because with almost 90% reduction in inflammation in the Phase II trial, it seems like a no-brainer that it should work in Phase III as well. If there's link? And then also, I just wanted to hear, Martin, when you say reasonably fast conduct of the trial, could you also try to give a little bit more color on that, knowing that, for example, CANTOS was 3.5 years in progress of what -- how fast will your trial be compared to that?

Great. Thank you, Carsten. I like the notion of no-brainers in R&D. I'll remember that for another set. So Marcus, I think it goes to you. So now Martin is going to spend significant resources with more than 6,000 patients in Phase III. So what makes us confident that the GIP reduction will translate into a MACE benefit? What's the evidence you have?

I think I'll make a start, Martin can chip in. But you highlighted rightly CANTOS right. And I think for us, it is a fundamental source of inflammation. And if you look into the categorization of the high CRP population there, I mean, they had a significant benefit on endpoints. And ultimately, I think that is the fundamental value proposition, number one. The other one is I think we have increasing evidence, so genetics, Mendelian randomization that puts IL-6 at the potential center of some of those effects, right? So I would argue it's circumstantial evidence because the point you're trying to test here is how does the reduction of hsCRP ultimately reduce cardiovascular risk. And ultimately, that is the hypothesis we are now testing. And maybe the second part of the hypothesis is, well, if we're blocking IL-6, can that be delivered in a safe way, right? Because as you know, obviously, IL-6 and through the IL-6 receptor also plays a role in sort of the immune response. And we're really trying to find a sweet spot, and we do firmly believe that ziltivekimab is actually the best drug to test this hypothesis, where we see very clear evidence for a reduction in hsCRP, but at least in the Phase II trial that we have seen, a very benign side effect profile. So when you put this together, I think we have a good drug modality now to take it into a larger outcome trial. Now CANTOS was a rather heterogeneous all-comers population. We're now zooming in into much modifying patient population. And by the way, I think that is fundamentally probably the way to think about some of those trials in the future that we are moving from all-comers to more specified and more stratified patient population, just as a general theme. Martin, anything you want to...

Yes, I very much agree. I think we have in a Phase II setting everything that we can have. We had to rely on biomarkers. This is a new setting. I sometimes use the analogy of originally when we started to talk about LDL lowering, someone had to do the outcomes trial to actually show that this was associated with improved outcomes. And this is specifically the situation where we have, to Marcus' point, very strong Phase II data, both from an efficacy, but also a very benign safety profile in this space and given the mode of action. This obviously does call for some risk on our part. So it's not the lowest risk Phase III trial that we've conducted so far. But we do believe that we have derisked it to the extent that we can given, again, the very good safety profile, the right patient population and a benign -- sorry, good efficacy profile. From a trial contract perspective, obviously, in these outcomes-driven trial, it very much depends on the expected event rate. I know, Carsten, you would love me to speculate and give you a very hard data. And obviously, I can't do that because this is also a new population for us. We made some assumptions. And I'll be happy to discuss in a couple of years if we made the right assumptions or not. That being said, even if we see a very high event rate, higher than we expect, we would still be required to expose patients for a certain number of years, probably at least 2 years for a proper safety event -- sorry, safety assessment on behalf of the regulators. So again, balancing the event rate with a requirement -- minimum requirement of exposure.

Thank you, Carsten. Thank you, Marcus. This concludes today's Novo Nordisk R&D investor event in connection with ADA. We thank you for listening in and asking a lot of really good and relevant questions. We will be back with our Q2 results release in August 5. So looking forward to connect at that point. In the interim, I wish you all a beautiful summer until then. Thank you.