Good evening, and a warm welcome to Novo Nordisk Investor event in connection with the American Heart Association annual conference here in Philadelphia. My name is Daniel, and I'm heading up Investor Relations at Novo Nordisk. Also, a warm welcome to those of you following the webcast online. I have to say that even though we're going to talk about SELECT, and the results have been published today, that we could still end up talking about forward-looking statements, so please look at the disclaimers as always. This is the agenda. We are going to discuss SELECT in detail, and we'll have two presenters, and I'll get back to them. Just before we dive into SELECT, this is our strategic aspirations. This is the way we report on our strategy progress in Novo Nordisk.
We recently reported our results for the first nine months of this year, and we continue to make progress on our different aspirations. Notice we now treat around 40 million people within diabetes, majority of those being on insulin, but of course, also more and more being treated with our innovative GLP-1 treatments. Within our commercial aspirations, we have now reached exactly one third of the diabetes market, and we have also reached more than DKK 25 billion in obesity care sales. But as you may recall, our aspiration says we want to do more than that, so we keep on pushing, and then we'll continue to do more.
We have been growing sales 33% in the first nine months of this year, and also a very strong growth in operating profit, and we raised the guidance for the full year, also here in earlier in October. Then the reason we're here today is, of course, to talk about our R&D and in particular our obesity aspiration, and that's what Martin and Robin are going to do in a minute. Martin Holst Lange will be starting out by talking about the results from SELECT. Martin, most of you know him, Executive Vice President for development, long career in Novo Nordisk, various positions, and the last three, four years, part of executive management and a key person in terms of running the clinical trials.
Robin Evers, many of you have also met Robin before. In a few weeks, Robin will have been 10 years with Novo Nordisk. Before that, he worked in, among other, Pfizer and Wyeth, and Robin is today heading up regulatory, safety, and quality assurance in Novo Nordisk. With that, Martin, I'll give the word to you.
Thank you very much, Daniel. Also, a very warm welcome from me. Obviously, most of us have been discussing SELECT for years now, and really, really great to be here today because, we've not been able to share data for the last couple of months. We've obviously known most of the data. We are still digging into parts of the sub-analysis. I'm receiving a lot of questions for sub-analysis where you guys are, I would say, almost more inventive than we are. And actually, you've given me some good ideas today, but most of what we share today is obviously something that is not only part of the original analysis plan, but also part of what has already been submitted to regulatory authorities in both U.S. and Europe.
Very briefly, and we all know this, cardiovascular disease is one of the leading causes of mortality on a global scale. Even today, even with what we see as being great advancements in treatment of dyslipidemia, treatment of hypertension, treatment of actually diabetes, we still see cardiovascular disease as a major driver of mortality. Big part of that mortality is associated with diabetes, and a little bit dependent on the study, it's estimated that approximately two million deaths every year is attributable to obesity. But that basically means that there's a tremendous unmet need. This is what we want to address in Novo Nordisk. Also, because we never saw treatment of obesity just as a weight loss exercise, this is about changing outcomes for patients for the better, and that's why we decided to do the SELECT trial.
17,500 patients, actually 17,604 patients, to be precise, 41 countries, more than 800 sites. Approximately five years of duration. This is event-driven, and that basically means that first patient in was treated for four years and eight months, but last patient in was treated for approximately two years. I come back to the average treatment period. The primary objective was obviously to look at the semaglutide 2.4 mg versus placebo in terms of its ability to lower the incidence of major adverse cardiovascular events or MACE. This obviously on top of standard of care, and standard of care in this context is primarily focused on treatment of cardiovascular disease or risk factors for cardiovascular disease, including obviously dyslipidemia and hypertension.
In addition to that, we had a number of secondary objectives that's also reflected in the endpoints. But obviously, what we want to look at, the cardiovascular risk factors, risk of mortality, both specifically CV mortality, but also all-cause mortality, glucose metabolism, body weight, renal function, and so on. The design was, in principle, very simple. Patients were randomized one to one to either semaglutide 2.4 mg or to placebo. As we just discussed, the study was event-driven, but the overall duration of the trial was four years and eight months. Primary endpoint, three-point MACE, that's myocardial infarction, it's stroke, and it's cardiovascular mortality. We had a series of confirmatory secondary endpoints, starting with CV death at top of the test hierarchy, heart failure, composite endpoint, and then obviously all-cause mortality.
In addition to that, we had a great number of secondary endpoints, all of which have on the cardiovascular side been disclosed today, but some of which we are still preserving for later publications and later communication. Patients were your standard obesity with cardiovascular risk population, age above 45, BMI above 27, established cardiovascular disease, meaning either prior myocardial infarction, prior stroke, or symptomatic peripheral artery disease. Patients suffering from diabetes were excluded. We had a long list of other exclusion criteria, but I think in this context, the most important was the exclusion of patients suffering from diabetes. And obviously, again, you know the broader trial information. I think it's maybe important here to call out that in opposition to our previous trials, patients and treating physicians were allowed to reduce the dose and still being compliant with the protocol.
Patient population, more or less, what you would expect in this space, so, more males than females. I'm oftentimes being asked about that, but for better or worse, males are more susceptible to cardiovascular disease than females. So even if we normally see more females than males in our obesity studies in a cardiovascular setting, this is largely what you would expect. Mean age of 62 years of age and mean body weight of approximately 100 kilos. That leads to a mean BMI, a baseline of 33, which, as you probably will recall, is substantially lower what we, than what we had in the STEP program. So some differences in terms of the baseline population as compared to the STEP program.
These were reasonably equally distributed, but approximately 30% of patients had a BMI between 27-30, which will be important when we look at the CV impact when based on BMI categories. Patients did not have diabetes, but approximately two-thirds had pre-diabetes. And that's obviously also interesting when thinking about the progression to diabetes during the trial. In a five-year trial, a number of patients will be expected to progress to diabetes. Of interest, based on the inclusion criteria, approximately 70% had prior myocardial infarction, 20% had prior stroke, and only 4% had a peripheral artery disease. Also, importantly, 25% had heart failure. So a lot of potential important information from that, as well as the fact that a large proportion of the patients had a degree of chronic kidney disease.
I would say virtually all our patients had dyslipidemia, and they were well treated for that dyslipidemia, chiefly through statins. Patients had hypertension, and they had a level of inflammation as measured by hs-CRP. You will also notice that level, if you have really dug into our previous publication, is lower than what we saw in the STEP program. So in the STEP program, hs-CRP baseline was approximately four. Here, it's approximately two. So a different population. I think by now, you know the primary endpoint. We see a very nice 20% reduction in terms of the primary endpoint, reducing risk of myocardial infarction, stroke, and cardiovascular death. What is obviously important to call out, and that has been called out several times during the day, that we see actually a divergence between the lines almost from day one.
That is in a cardiovascular outcome setting, a little bit unusual, but it's probably speaking to the impact that semaglutide does to these patients or has to these patients, that to an extent, we see as independent or at least partially independent from the weight loss. It was a well-conducted trial. Almost 97% of patients completed the trial, of which approximately 75% in each treatment arm completed the trial on drug. And then obviously, there's a dialogue, is that good or bad? In a five-year trial, I can tell you it's very good to have 75% of all patients not only being on drug in the active arm, but actually also being on placebo in the placebo-controlled arm. Mean exposure time was 33 months in the semaglutide arm and 35 in the placebo arm.
We've had a lot of questions, both today and before today, on different categories of cuts, not so much on age, but as you will obviously see, maybe. As you will obviously see, it doesn't really matter in this space, which age you have. There's no real difference between the different age groups in terms of the cardiovascular benefit that we see. More importantly, I think this is a really, really nice for us discussion point, is that it doesn't really matter which BMI you have. There has been a lot of theories going into this, that with higher BMI, you would see higher benefit. That's actually not the case. You see the same benefit throughout. I've also received some questions if the benefit is less at the higher BMI groups.
You'll see that the confidence interval here is very broad, and that's basically because we're talking very few patients and very few events at those levels. I don't think we should infer a lot from that. So our conclusion is really no difference in terms of the CV benefit, regardless of BMI category. We can also look at patients who—this is really annoying, and I don't think that you've seen this before. If you look at patients who had BMI versus previous stroke versus peripheral artery disease, it's clear that the patients who came with a history of myocardial infarction has more benefit than the least patients coming in with a stroke.
In terms of heart failure, there's an indication that you see benefit in both, but somewhat more benefit in patients who did not have heart failure at baseline. From a renal perspective, it doesn't really matter. You see a profound benefit regardless of renal status at baseline, and basically the same thing if you look at hemoglobin A1C at baseline, above or below 5.7%. So very robust and very consistent data. I don't think, at least I've been doing this for 22 years, I've never seen consistency like this across everything that we look at. You've seen the first slide, the cardiovascular death. We have a 15% non-statistical difference between semaglutide and placebo. We've been asked throughout the day on the shape of the slide. We believe that the shape is caused by COVID-19.
So patients are still dying, but during COVID-19, the period, of COVID-19, a lot of deaths were ascribed to dying from infection rather than dying from cardiovascular disease. So, and I'll show you this in just a minute. If you look at the all-cause mortality, it's more a straight line, and we believe that the explanation is basically patients are still dying, but being ascribed—the mortality is being ascribed during COVID-19 to dying from infections rather than cardiovascular disease. Doesn't really matter. At the end of the day, we see a 15% difference, and again, the slide, the trial was not powered to look at difference on the subcomponents of the three-point MACE. So actually getting close to statistical significance from our perspective is a really gratifying observation.
Even more gratifying is obviously to see highly statistically significant improvements in myocardial infarction, 28% reduction, and very statistically significant using the five percent significance level. But because cardiovascular death was not significant, we cannot test for multiplicity and obviously not claim superiority for myocardial infarction in this space. And then Time to non-fatal stroke , we see a 7% difference. This is obviously far from being statistically significant, and while all are pointing in the right direction in favor of semaglutide, I think it's fair to say that the benefit is primarily driven by cardiovascular death and myocardial infarction. If we dig a little more into this, we see the composite heart failure, which is statistically significant, again, using the five percent significance level, but again, not testing for superiority because of the testing hierarchy.
And similarly for heart failure, which is obviously incredibly exciting for us. All-cause mortality. We see a 19% reduction in all-cause mortality. This is again, statistically significant, and will be very informative. And going a little bit back to my explanation on cardiovascular mortality, here you actually see two reasonably straight lines, suggestive of patients are, for better or worse, still dying more on placebo, than on semaglutide, but during COVID-19, more ascribed to death from, COVID or infection rather than, cardiovascular death.
So this is, I, I'm not supposed to have favorites in what I do, but this is actually one of my favorite slides in my career, because, again, very rarely do we see assessment of a multiple of cardiovascular endpoints, and every single assessment is directionally going in favor of semaglutide, regardless of how we look at these data. Talks to the consistency of what semaglutide does to the cardiovascular system and the broader benefit that we see with this intervention. So we conclude if just focusing on the cardiovascular system, that obviously we meet the primary endpoint, 20% risk reduction in risk of having a myocardial infarction, stroke, or cardiovascular death. We do see these benefits occur very early, and they are directionally consistent across all cardiovascular endpoints that we have been looking at.
We do not see superiority for cardiovascular death, and therefore, we do not follow the testing hierarchy to the end. But we do see significant reductions, importantly, obviously, for heart failure, but even more importantly, for all-cause mortality. Both reaching the 5% significance level, but not claiming superiority as they didn't meet the test hierarchy. We've also discussed body weights, and I've had some questions, why, why do we see 10% body weight rather than 15% body weight? I'll just remind you, this is a different population, so more males than females, lower baseline BMI. It is a study where patients were not pushed to titrate their dose. They were not pushed to be adherent.
You saw maybe some data today, very, very briefly, that 70% of patients were at 2.4 mg throughout, which is actually substantially lower than what we see in our weight loss-focused trials. Then obviously, the ability to, for, for cardiologists to focus on titrating based on weight, that was not in their focus, and therefore, we did not pursue this. We have been asked, why did we not diligently go after the maximum weight loss? Our purpose was to do something that is close to real life, so this can be relevant for treating physicians, and I think we achieved that. What is incredibly gratifying for me is obviously, that this is the first time I see all your data on semaglutide, and the line is flat.
Some of you have even said it, it actually continues to go down. I will choose to call it flat and stable throughout the four years, and that is obviously speaking to the consistency of what we can achieve with semaglutide, not only in terms of the cardiovascular benefit, but also on the weight loss benefit. If we go into a little bit more detail, we have been discussing the risk of developing diabetes. In an obese population, that risk is there, and if I can actually see the slide, if we start with the placebo arm, you will actually see at baseline in both arms, approximately 2/3 of patients have prediabetes. In placebo, that number over four years increases to almost 80%.
A quite large number of people move into the diabetes space. What is also interesting is, for semaglutide, it goes the other way. And that's basically a testament, obviously, to what semaglutide can do on treatment. What is also interesting is what's gonna happen when we stop treatment. So as you know, we do the SELECT Life, and it will be interesting to see if that benefit that we've demonstrated here is something that we can achieve broadly and sustained. So what we see is a 73% risk reduction of developing pre-diabetes in patients who were normoglycemic at baseline. And we see an approximately 70% reduction in patients who have had pre-diabetes in terms of developing diabetes. Just want to spend some time because a lot of you have asked about what is driving the CV benefit.
The short answer is we don't fully know, but what we do know is that it is more than just the weight loss. You saw that indication from the fact that we saw a split of the lines already very early on. That's indicating that we see the CV benefit before we actually see the weight loss. And that also is what we can see when we do a mediation analysis. And in that mediation analysis, factors that matter are body weight, its glycemic control, its inflammation, its dyslipidemia, and its hypertension. It's also something that, we, for lack of better, call the semaglutide X factor. But, we focus on the things that we can see right now, and that is obviously, we know semaglutide improves blood pressure. We know semaglutide improves dyslipidemia, specifically, when we look at triglycerides.
I think importantly, we know that semaglutide reduces markers of inflammation, in this case, by approximately 40%. All of those will go into driving the CV benefit together with the weight loss. I think that is incredibly important. I've also been asked, when will you see the mediation analysis? We can't share it because we have external investigators who own the right to publish, and obviously they want to publish that before we share it broadly. So I think with that, I will leave it to Robin for safety and next steps.
Thank you very much. Thank you, Martin. And good evening to everybody. Good evening to people online as well. I'm gonna talk you through now just a few of the information related to the safety outcomes from this study. This is our largest-ever study of semaglutide conducted in a randomized setting. It's actually now the largest-ever study of any GLP-1 conducted as well, and it provides for us a unique picture on some important questions that we continue to follow, events of special interest related to semaglutide in particular. The study was robust, as Martin's outlined, but it gave us a lot of information related to serious adverse events.
There was a special collection of adverse event information in this study in which we collected serious adverse events, events of special interest, adjudicated events, and then events related to thyroid cancer, suicidal ideation, and any events related to discontinuation. So there was a special safety collection procedure in this study, which related to how we collected events and the number of events that we collected. We also obviously had a large number of patient visits and over six million lab assessments as well, and we continue to evaluate all of the information from that. What we see then in terms of the overall serious adverse events is that there were fewer serious adverse events on semaglutide compared to the placebo arm. So an incredibly reassuring outcome in the very highest level of what we see.
When you look at the breakdown of those, it's relatively well-represented across the different categories of events that we've seen before. But obviously, we see less cardiac disorder events, adverse events, serious adverse events, and less infections and infestations, and that may also then correlate to what Martin explained about in the related to the potential during the reporting of of death in during the COVID era as well. Other events are well-balanced across arms. We do see a slight significant difference in terms of surgical and medical procedures as well, but relatively well-balanced across arms from the other events there on the slide. Related to unexpected findings, there were none.
The remarkable picture for us, and as we look forward to communicate that to the medical community and in labeling and with the regulators, is a very reassuring picture overall for semaglutide. So we continue to see in the events of special interests no real difference in terms of malignant neoplasms. This is an area that we've followed for a number of years now with respect to semaglutide, and obviously, this is the longest exposure in a single study. Gallbladder related disorders are slightly elevated, something again that we've seen in other studies with weight loss. Acute renal failure, though, and acute pancreatitis, very reassuring, no difference between treatment arms.
The SELECT safety group also then evaluated two areas that continue to be of interest to us and also, continue to be of interest in our discussion with the regulators and ongoing assessments. This data set provides a very reassuring picture on those particular two areas of interest with the regulators that we, we have follow-up with. Firstly, talking about malignant neoplasms. We've talked about the overall rate, first of all, in terms of comparability between the two arms. But when you look at MTC and thyroid cancer, again, no real difference between the two treatment groups. Incredibly reassuring outcomes on the safety, safety collection from this study. And then when you look at psychiatric disorders, and obviously, we've been asked questions around suicidal ideation, again, complete equivalence between the two treatment arms that we see here.
When you look at the adjudicated events, again, no real difference between the two treatment arms as well. So overall, in terms of picture of events that are of special interest with the regulators and for us to continue to collect data, very reassuring data out of a large data set, now the largest that we have for any GLP-1. Then when you look at permanent discontinuations, you obviously see a higher rate of permanent discontinuations as of adverse events, serious adverse events or adverse events, sorry, not serious, adverse events leading to discontinuation. Here you see a larger proportion of adverse events in the semaglutide treatment group, and most of that's driven by the gastrointestinal disorders, something which is very well characterized for the semaglutide treatment group or GLP-1s overall.
We're no different than any of the other GLP-1 groups here, and no real difference between any of the other arms or nothing of any concern here in terms of patient discontinuation, adverse events driving patient discontinuation. So mainly driven by gastrointestinal events here. So overall, the safety picture, very well tolerated, well tolerated over a large period of time, median duration explained by Martin in terms of the picture there. You know, the picture there that we see mainly in terms of the serious adverse events, lower in the semaglutide treatment arm overall, compared to the placebo arm, and any difference is driven mainly by infections, surgical procedures, and the CV disorders events.
In the events of special interest, no real treatment difference between the groups, slightly elevated in the gallbladder-related disorders. Again, something that we've seen in the other studies as well, often related to the weight loss that has been driven. And then in the SAE, AEs driving discontinuation, slightly higher rates, mainly driven by GI adverse events driving that. So very reproducible from our STEP and studies overall and our previous SUSTAIN studies in diabetes, a very reassuring picture that we've presented now, both today but also to the regulators, and will be part of the labeling, obviously, when it's integrated there. Next steps. Put the slide out on next steps. Okay, one more. Okay, so what are we doing with this? As you sorry.
As you've understood from us, we've already made regulatory submissions now to the FDA. It's actually been given a priority review by the FDA, given the nature of the outcome of the study. So it's under a priority review. It's also been filed with the EMA, and we're continuing to pursue the regulatory submissions globally around the world now on the background of the primary outcome and the regulatory submissions that have been put together. We believe that the outcome of the study was... Well, it was designed with the regulator's input from the very beginning, so the robust outcome that we now have, I think, provides a lot of reassurance as we go into the end game now in terms of being able to get this into labeling and into future promotion of the products.
We believe the strong outcomes from the primary endpoint analysis is really driving how we see the positioning of the product into the future, and obviously, we continue to present all of the efficacy and safety information in a robust way for the regulators. And obviously, it becomes a totality of assessment on the basis of the data for the final approval. SELECT Life is now ongoing. It's due to follow up for 10 years, six-monthly interval. The follow-up for patients that participated in this study will be capturing events in a patient-reported outcome measure in terms of contact with those patients on a periodic basis. And in terms of the final closure, Martin, I think this is gonna be for you.
Thank you, sir. So this is the end or the beginning? I don't really know how to phrase it. Obviously, SELECT is done. We have been on this journey for the past six years, so we are incredibly happy to be able to see a 20% risk reduction in terms of our primary endpoint MACE, but also consistent beneficial findings in all cardiovascular assessments that we've done in addition to that, as well as improvement in renal function, lower risk of developing diabetes, and a very nice and attractive safety profile that is in line with what we expected, but also in line with what we previously have seen with GLP-1 analogs. So we obviously look forward to regulatory interactions, interactions with treating physicians and patients, but certainly also with interactions with payers.
It's an incredible tool to have SELECT when we discuss, when we negotiate with payers and decision makers in years to come, because obviously they will ask: "Is this just about weight loss? Is this just about throwing off some, some pounds?" And the clear answer is no. This is about improving outcomes. And we've shown with semaglutide that it actually matters that we have the right intervention. So thank you for your attention, and I think now it's time for Q&A.
It is now time for Q&A, and, let's go with one question per person, and, please state your name and institution, and then maybe we can do several rounds if we have time. Let's start over here with Richard.
Hi. Thanks, Richard Vosser from JP Morgan. So we've seen the correlation of benefit of BMI, but what's the correlation of benefit with weight loss? Does it follow the BMI? So the lower BMI maybe have lower weight loss and therefore it's the same, and so it's the same across lower weight loss. So thinking about the, you know, a 5% or a 10% weight loss, whatever cut-off, how does it follow?
So we've done that analysis... Sorry, Danny is moderating.
You go for it, Martin, honestly.
Good. I need, I need permission.
Sometimes you can't hold him back, honestly.
No, exactly.
Can you imagine what it's been like the last two months? I tell you.
All right, we've done the analysis, and while obviously it's a post-hoc analysis, what we can see from a statistical perspective, baseline body weight and weight loss is not a key driver of the cardiovascular outcomes, basically in extension what we also saw with the BMI correlates.
Nice and clear, Martin. Thanks for that. Let's go to Emily, next to Richard.
Hi. Thanks. Emily Field from Barclays. Just one question, we've been getting asked today just about the results in the North America geographic subgroup, why that maybe didn't reach statistical significance. And then, I guess, just more of a holistic question, just, you know, how are the learnings from SELECT impacting how you're thinking about designing further trials for CagriSema, given that we're kind of just relatively just around the corner from seeing that first kind of the obesity data next year, hopefully?
So let's take the first question, and maybe we'll get back to your second question later, Emily. But Martin, the first question?
Yeah. So first of all, you can never, in any sub-analysis, expect to see statistical significance, just like the primary endpoint. We're super happy to see the a significant difference for myocardial infarction, but you shouldn't expect that. The U.S. population was 20% of the entire population, so to see something significant there would be a little bit unrealistic. The point estimate was a little bit higher on the primary endpoint, and that is a reflection of a little bit that uncertainty.... but it's also not unusual to see that in a U.S. population in a cardiovascular outcome trial. So not really something that concerns us. And specifically for CagriSema, what we know with CagriSema is that on every cardiovascular risk factor we look at, CagriSema appears to work even better than semaglutide in monotherapy.
Obviously, we think that when we think about not only cardiovascular, the data that we want to generate, but also in other indications.
Thank you, Martin. To Michael.
Thank you. Michael, try to speculate a bit about what is actually driving this very early, early separation, because it's been a major topic today. And I guess at some point in time, we go to pay this. It's also like an episode of the last two days, we don't really know. So, a bit more speculation on at least sort of the most significant drivers of that.
So, I mentioned the five key drivers, and if I try to rank them, it's actually weight loss, inflammation almost at a tie. And then obviously, glycemic control, and then dyslipidemia and blood pressure. But again, we're not giving specifics at this point in time.
Maybe later, improved renal function.
Yeah. Yeah.
Yeah.
Sorry?
Maybe later also, improved renal function.
Good. Thanks for that. Let's move over here to Pete.
Thank you.
Citi. One question for Robin or Martin, one of us in, feel free, but regulatory strategy. After the second session today, speaking to some of the KOLs, you know, they say it's gonna be-- the pressure on Medicare is gonna grow pretty- it's only a matter of time until the New York Times starts writing articles saying, "Why is Medicare patients not getting access to this drug?" So I know you can't go for an obesity indication, but can we at least speculate with you about how aggressively you'll try and pursue CVD indication, or persuade CMS to provide coverage on the CVD side? Martin?
Yeah, so I actually don't think that's a regulatory question.
No, it's not.
... because anything Medicare Part D would require changing of the U.S. legislation, and we're not speculating in that. The technicality that you're thinking about is to go via non-obesity approvals. I don't think that we see that in the cards right now.
Good. Let's take, Seamus.
Oh, thanks. So the speculation on the closing of the curves, can you just talk a little bit about the cardiovascular death versus the all-cause mortality? All-cause mortality is amazing. You've explained that, but have you extracted the COVID infection data, so 24%, 23% of the patients, and in extracting that, has that actually become apparent?
Yeah. So it's a really good question. You saw some indications. Sorry. I have to learn. We saw some indications from Robin's slides, where you actually saw that infections and infestations, serious adverse events, lower with semaglutide than with placebo. You didn't see the mortality data, but that would be the same thing, so lower with semaglutide than with placebo. Specifically for COVID, you also saw exactly the same number of patients had COVID during the trial. What semaglutide does was to protect you from serious adverse events and mortality. And this is with a big caveat, and Daniel is gonna kill me afterwards because it's my calculation and not based on a statistical calculation.
But when I look at the data, the deaths from COVID-19 is approximately 30% lower with semaglutide than it is placebo. But it's not a formal statistical calculation. It's Martin Lange.
This was the forward-looking statement I talked about.
Yeah, exactly.
We'll go to you.
Thank you, Martin Parkhøi from SEB, and back to Michael's question on the secret sauce in semaglutide. Why do you think that the X factor call should only be related to semaglutide and not all GLP-1s? If you look at the albiglutide study on the Harmony study, that was up to 28 weeks, there was no body weight difference, yet there was a 22% CV benefit driven by MI. So it... Why should it be different for semaglutide?
It's a good point, but it's exactly that that we are speculating in. I mean, for better or worse, albiglutide is not on the market right now. And that basically means that if you look at the GLP-1s that are currently on the market, semaglutide, at least in the diabetes space, and diabetes is the only place where we have data from other GLP-1s. At least in the diabetes space, stands out in terms of the magnitude of CV benefits. So most GLP-1s have CV benefit. You also saw that from the slide today. But semaglutide is 20% risk reduction in Type 2 diabetes. The next one is 13. So the magnitude of CV benefit is differentiated, it appears, between GLP-1s.
What allows semaglutide to do 26%, it's—I think it's a combination of a little bit the X factor, the body weight loss, the improved glycemic control, the inflammation, and so on. But again, in type 2 diabetes, we see differentiation in the magnitude-
The renal effect.
Sorry, just to-
Yeah.
... Let's go to Pete.
This might not go straight to Martin, but I know, apologies for this, it might be the right person in the room. But just thinking bigger picture from a commercial standpoint, is the thinking in Novo, is it now you go to cardiologists? So I think you said but the main population who enrolled patients in this study, I mean, correct me if I'm wrong, and detail this as a cardiovascular drug for the cardiologist patient population, rather than perhaps, I guess, currently the endocrinologists and primary care doctors that you're more often targeting.
Martin, we give it up.
You're absolutely right, I'm not the commercial person. So, again, a little bit of a forward-looking statement, but I think it's fair to say, and I think you also heard that from Steve today. The way that we think about this is, this is still patients with overweight and obesity and cardiovascular disease, and that's our starting point. But a lot of these patients are being treated by cardiologists, and our approach will then be to go to cardiologists and say, "If you have patients who are overweight, obese, with established cardiovascular disease, then we have a good treatment option for you.
I think it completely follows everything that we've heard today about the revised AHA guidance practices about how cardiologists look at the disease in terms of cardiology, kidney, and the metabolic syndrome. You see? That's how cardiologists are looking at the disease.
Good. Let's go to Mike.
Thanks, Mike Nedelcovych, TD Cowen. Now that SELECT is complete and SELECT Life has kicked off, when do you expect you'll have the diabetes prevention data that you need to convince regulators?
Yeah, I think I'll take the first bit. I think in terms of... I think the regulators are continuing to evolve, particularly the FDA, is thinking around what will be needed for a pre-diabetes or a diabetic prevention claim. They certainly will be impressed, I think, by the data that we'll be sharing them in the context of the SELECT outcomes and what we've shown in terms of returning patients to more normal glycemia from a pre-diabetes state. I think what the regulators, and particularly the FDA, have talked about is a long period of follow-up, or period of follow-up off drug, and to see whether or not there's a reversal of a patient population off drug.
Obviously, we'll continue to follow the outcomes from this study, but we'll put that together with other data sets that we'll be pulling together. It's not exactly clear yet what the FDA will require prospectively to be able to make that claim.
Good. So, Michael first.
Just for Robin on, on heart failure. So how should we think about sort of the, the overall regulatory submission? One is also, we need the data from the diabetes part, but will be sort of a pool analysis of, STEP and HFpEF and, and SELECT, that is sort of needed to, to get the claim in, in heart failure.
Yeah. So I think at the moment, shouldn't look at one individual data point. I think the data point from SELECT provides really reassuring and positive outcomes relative to heart failure. But we have a number of studies underway in terms of heart failure that we'll be assessing and reporting on, and have reported on already in terms of pulling the data sets together from a multiplicity or multiple data sources, and making sure that there's corroborating evidence from multiple data sources that provide reassurance to the agency in terms of the outcomes. And with nothing more definitive to say now, but we will be taking those discussions forward.
We'll go to Richard.
Hi, thanks. Just thinking back to the renal endpoint, it's a composite endpoint, so how much of that composite do the actual renal events contribute? And is it, sort of, congruent with the CV benefit? So are we seeing the same benefit in just renal events in that?
From a numerical perspective?
Go ahead.
Thank you. You heard the answer from Steve today. They don't want us to go into too much details, but I think it's fair to say that we see a renal specific benefit that is, from a magnitude perspective, may be comparable to what we see in terms of the cardiovascular benefit. I think you get that indication, albeit in a Type 2 diabetes population, from FLOW, where, as you know, we powered the study for 20% difference, and the interim analysis and the DMC would not allow us to stop unless we were on the very good side of 20%.
Thank you, Martin. We go back to Mike from Cowen.
Thanks again. You've been asked many times at this point, what the average treatment duration on Wegovy will look like in practice. We now have a number, at least from a long-duration trial of around 33 months. Is that as good a guess as any for what average treatment durations will look like in practice, or should we have a very different number in mind?
I think it... Sorry.
No, please go on.
If we look at the SELECT data and if we accept that it is a proxy for real world, for better or worse, that was a very low intervention in that study. Having 75% of patients on drug and on dose after 4 years of treatment, that's maybe a reasonable number. I think what we are seeing so far, and again, that's early days, and we're not committed to give you in-market data until next year. I don't think it's unrealistic, but that is very concretely a forward-looking statement.
Thank you, Martin. Let's go back over to Seamus.
... Thanks. So if you were going to adjust anything at all in the CagriSema outcome study, what would you adjust to ensure that you capture the entire hierarchy?
Martin, that's also for you.
So I'm not sure I would change the hierarchy. You have to make the hierarchy based on clinical relevance, but also in terms of what is it possible to get into, for example, a label. And from that perspective, I think Robin will agree, the bar is lower for cardiovascular death than it is for all-cause mortality for various reasons. So probably, or with some likelihood, you may see the hierarchy be the same. But obviously we have to discuss that, and we have to really dig into the SELECT data. That being said, I would probably like to see a, and I think that was also asked from stage today, a better gender distribution than what we saw in SELECT.
Would you power cardiovascular death differently, though? Would you ensure that you have enough events so that you get a definitive outcome because the study was not powered for cardiovascular?
That's an incredibly difficult question to ask because the actual power for detecting the 15% that was always actually quite low. So even if we assume that, it would require a very big study, and it obviously also depends on what can semaglutide actually do. Can it do more than 15%? Sorry, CagriSema, and obviously at this point, we don't know. But if you really wanted to power for the subcomponents, it would be a very big study.
Thank you, Martin. We'll go to Jasper.
Thank you. Just on the topic of health economic modeling and the cost for a CV event. So when you do the calculation, it seems like it costs $1 million to prevent a CV event. What is Novo's answer to this, and how will you use this data to sort of convince people on the health economic modeling and benefits?
Martin, we count them. Yeah, number needed to treat.
Yeah. So I think from our perspective, first of all, I'm not exactly sure your numbers are correct, but we can discuss that. I haven't seen a full health economic model. But I think from our perspective, this is not just about the primary input. It would be incredibly simplistic just to look at that. And of those, actually, from a pure health economic perspective, myocardial infarction and stroke would be the key drivers of the primary endpoint. But you should also be looking at, for example, heart failure, which is incredibly expensive, kidney disease, incredibly expensive, risk of developing diabetes. And what we haven't shown today is that we've actually also in fact been looking at risk of being hospitalized, proportion of patients being hospitalized, number of days in hospital, and so on.
All of that will go into the healthcare model. I think it's really, really important to keep in mind, it's not just about the primary endpoint from a health economic perspective.
Thank you, Martin. We'll have a question there, Marcus Schindler.
Thank you. On the five factors you feel are important to drive the benefit, the one we have in similar to STEP is CRP. So going back to your point on SURPASS-4 and SELECT and SUSTAIN-6, is that the sort of key missing piece where you believe that semaglutide is superior? And just on a related matter, have you, with the oral GLP-1 receptor agonists, so the small molecules, on your five factors, how do they stack up versus a peptide? Are they also inferior on inflammation based around CRP and proBNP?
The smart way of sneaking in two questions, but it was your first time, Mark. So Mark, can you give short answers?
So, I haven't seen a lot of data on CRP from Lilly. I actually saw some today. They also have a reduction in CRP, so I don't want to speculate exactly what is driving the difference in diabetes. I think on the small molecule side, we haven't seen data. So, I don't think I can speculate on that.
Thank you, Martin. Martin, ...
Yes, just a question on the difference between SUSTAIN and SELECT we went back to. Do you know why we see the impact on MI in here and on the SUSTAIN trial, it was on stroke? Is there something due to the link between diabetes and stroke, which is more evident than the link between stroke and obesity, that you see the difference?
Martin?
So at this point, and obviously we'll look into this in more detail, but at this point, your guess is as good as mine. It could be the difference between diabetes and obesity. It could be from a geographical perspective, differences in certain people. So honestly speaking, I don't want to speculate.
Let's take Pete next to Martin.
Thanks, Pete. One question on the incremental data from SELECT for next year. So Professor Deanfield called out the ACC presentation of the heart failure data and gave us a teaser that we worth being there for. It sounds like the renal data will be presented. At this stage, can you say anything else about what other endpoint, what other data were presented, and can I just squeeze in a quick second question? Have you looked at the patients that were on drug for four years? I think last when I saw you in London, Martin, you said, I think 20% of the population would have been on the drug for over four years. Have you had a chance to look at that subgroup, and, you know, is there any difference in outcome?
So the short answer to the last question is no, we haven't—at least I haven't seen that analysis yet. And to your first question, I think you will see us appearing with SELECT data at a number of conferences next year, including obviously also American Diabetes Association.
Further questions we still have time for. Let's take Richard Vosser on the mic, and then we may come back over.
I just wanted to come back to the surgical intervention or the medical intervention on the serious adverse events, and I think you sort of touched on it in terms of isn't that a huge driver of cost? So what can you use in a serious adverse event context on that front with the payers?
Robin, Martin?
Yeah, I mean, I think the data isn't in terms of claim, but in terms of totality of what the healthcare costs were, and we do have information in terms of total days of hospitalization as well that's reassuring there, so that will come into the economic model.
Or everything that can be quantified will go into the model where it makes sense. I mean, a lot of these models are using sort of predefined set points, but I think to your point, surgical procedures, it does make a huge difference. And the one I've seen is actually cardiovascular procedures. And I think it makes sense given what we are looking at.
Thank you, Martin. Rob, we'll go to Michael.
Thank you. So a quick follow-up on the diabetes prevention question. Martin, I think you've said in the past that your expectation was that FDA would ask for three or six months of data off therapy.
I think I've said 12 months, but yeah.
Okay. Is that still the expectation?
To Robin's point, we don't know. We have to have that dialogue with the FDA, but in recent or previous dialogues, they've mentioned 12 months.
Durable, I think, is the word they use.
Yeah.
But we don't know yet.
Good. We have time for a few more questions, then we'll go to Martin Parkhøi.
Thank you. Martin Parkhøi, SEB. Just because I was wondering, you saw a little bit change on the line after 36 weeks on the weight line, suddenly a drop, which exactly happens at the same time as you start to see a separation, higher separation on the CV. So are there any link to the certain change we suddenly see the weight drop and also the impact on the CV?
So, the impact you see on CV is, I think again, not a function of all of a sudden something happens, but actually that more patients are being, or more deaths are being ascribed to cardiovascular deaths rather than, let's say, COVID-related deaths. And, I think it's gonna be a little bit speculative to go further than that. Sorry.
Thank you, Martin. We'll go to Seamus.
Just a quick question. When will we see an on-treatment analysis of the study? And would that on treatment analysis basically come with cuts of the dose level, and the impact on the different endpoints?
So I think you will see that during the course of 2024, in different venues, but also different publications. Based on what we've seen so far, and we haven't done all of the analysis, it's not gonna change the conclusions, fundamentally.
Good. Any final question before we wrap up? It doesn't seem to be the case. Then, Martin, I don't know if you have any final words before we close the call.
I think we've said what is required to be said. This is, for us, the end of the SELECT journey. It's been a great journey. A lot of you have followed with interest and obviously looking at the outcomes, not only being able to demonstrate fundamental efficacy but also a really attractive safety profile for Wegovy is really great for us. But it's also a start of a journey. We see semaglutide, we see Wegovy, we see SELECT being able to change how we see obesity as a disease, removing stigma around obesity and how we will treat obesity moving forward. I think that is a benefit for a great number of people across the globe.
Thank you so much, Martin. This concludes our conference call here in connection with the American Heart Association. Thanks so much for attending the event and following the webcast, and be back in the end of January with our full year results.