Amgen Inc. (AMGN) Earnings Call Transcript & Summary

My name is [ Jay Son ], and I will be your conference facilitator today for Amgen's Analyst and Investor Conference Call at the 2022 American Heart Association Meeting. [Operator Instructions] I would now like to introduce David Reese, Executive Vice President of Research and Development. Mr. Reese, you may now begin.

Thanks, [ Jay Son ], and good afternoon and good evening, everyone. Thanks for joining us today for an update on our general medicine portfolio, selected molecules in development. We have a lot of material to cover today, some interesting new data, and so we'll move right into the program. On the next slide, you can see our standard safe harbor statement. And on the following slide, you see the agenda for today's call. I'll provide a very brief introduction with some context setting. And then, Professor Marc Sabatine, who's a Professor of Medicine at Harvard Medical School, Chair in Cardiovascular Medicine at the Brigham and Women's Hospital and the Chair of the TIMI Study Group, will review 3 sets of data updates from the FOURIER study and the FOURIER open-label extension studies with Repatha and then finally, our just published Phase IIb study of Olpasiran, which targets Lp(a) for atherosclerotic cardiovascular disease. Following Dr. Sabatine's remarks, Narimon Honarpour, who is the Head of Clinical Development for Cardiometabolic Disease at Amgen, will provide a short update on the AMG 133 program, we know there is a lot of interest in that, make some concluding remarks. And then we'll move into a question-and-answer session. Next slide. On this slide, you see a depiction of, contextually, one way that we think about cardiovascular disease. Among the universe of individuals with high cardiovascular or metabolic risk, and many times those overlap, there are various drivers of cardiovascular disease, in particular, and for today's discussion, atherosclerotic cardiovascular disease. Of course, LDL is the primary driver and should always be controlled. We'll provide updates, as I noted, on the Repatha program. But in addition, beyond LDL, there are other drivers of residual risk. One of the primary drivers in the number of patients is lipoprotein (a), and that is targeted by Olpasiran. As I think you'll agree, we've shown profound reductions in Lp(a) levels with Olpasiran. And Dr. Sabatine will provide a perspective on those data. One of the other drivers of cardiovascular disease and risk is obesity and its many downstream consequences. And as most of you are aware, we have a molecule in Phase I development. Full data will be presented in about a month for AMG 133, which is a molecule that antagonizes the GIPR receptor and agonizes GLP-1. And again, Narimon will provide a perspective on that molecule. So with that introduction, let me turn things over to Professor Sabatine to review some of the data that were just presented within the last 48 hours here in Chicago. Professor Sabatine?

Dave, for some reason, the room in Chicago, we can't seem to hear them. So they're checking into that right now to see exactly what's going on.

Okay. So if we can ask everyone online to be patient for just one moment.

Yes, just stand by everybody, please. [Technical Difficulty] Okay. So for some reason, the folks in Chicago, we can't hear them who are in the conference room in Chicago, so we need to get them connected ASAP. [Technical Difficulty]

Ladies and gentlemen, we have lost connection with the speaker line. Please stand by while we try to reconnect them. [Technical Difficulty]

All right. Okay. So let's see if we can hear the conference room in Chicago.

Hello?

Yes, Narimon, can you hear us?

This is Narimon. Are you able to...

Okay, go ahead. Yes, we can hear you fine now. Yes. Okay, Narimon. So I think we can go ahead and begin. Narimon, can you still hear us? Or did we lose you again? [Technical Difficulty]

The team should be back -- the team is back in now.

Okay. Narimon, can you hear us?

Yes. This is Narimon. We can hear you loud and clear.

All right. That's great. Okay, so why don't you go ahead and proceed? Thank you.

Yes. Great. Thanks, everyone. So this is Marc Sabatine here. Happy to share the data. So can we have the next slide, please? Great. So as I think many on this call will know, in the FOURIER trial, over 27,000 patients with stable atherosclerotic cardiovascular disease were randomized to evolocumab versus placebo. And we showed that evolocumab substantially reduced the risk of major adverse cardiovascular events. But there was no observed effect on cardiovascular mortality. However, that has to be interpreted in the context of the relatively short median follow-up for the trial, which was only 2.2 years. I'll remind people on the call that the pivotal statin trials had a median follow-up of 4 to 5 years. And those trials demonstrated both a lag effect. And by that, I mean, that the clinical benefit grew over time, and that was particularly true for the effects on fatal events, where those event curves did not diverge typically for about 2 years. And so therefore, we weren't really well positioned in the parent FOURIER trial to discern any effect on fatal outcomes. The trials also showed a legacy effect. And by that, I mean, that the clinical benefit persisted an extended follow-up even after the parent trial has ended. And that was seen in trials regardless of whether patients, once they were done with the trial, didn't start on statin therapy or all started on statin therapy, the issue with -- that the LDL cholesterols converged. But still, there were differences between the 2 arms based on their original treatment, and therefore, very long-term data on the safety and the efficacy of LDL cholesterol lowering, with PCSK9 inhibition are needed. Next slide, please. And so this shows the study schema. I already commented on the patients who we enrolled in FOURIER. And as I note here, the median follow-up was 2.2 years. And as is noted here, then a subset of patients, a little over 6,600 then entered into an open-label extension, where all patients got evolocumab. And importantly there, the median follow-up was 5 years, and stitching the 2 together, then the median follow-up was about 7.5 years. And some individuals went out to about 8.5 years of follow-up. Next slide, please. And so a couple of then points to share for these open-label extension data. First here, I show the effects on LDL cholesterol. You can see on the left-hand part during the parent FOURIER trial, obviously, the differences between the evolocumab and placebo arms. And then once the placebo patients transition to evolocumab, you see the robust reduction in LDL cholesterol, again on the order of 60%, 65% reduction in LDL cholesterol, then down to an achieved level of around 30 milligrams per deciliter, really one of the very lowest levels of achieved cholesterol in any major cardiovascular outcomes trial. This slide also nicely shows the durable effect of evolocumab. So now again, stitching together the 2 trials, you have about 7.5 to 8 years of follow-up, which I think is the longest experience with any PCSK9 inhibitor. Next slide, please. Here showing safety. And so just to orient you, we're looking at a variety of different safety outcomes that have been examined in this field. The blue and the red bars show what the annualized incident rates were in the placebo phase of the FOURIER and evolocumab. And then the dark crimson third bar there then shows, for the extension phase, for patients who were then on evolocumab for 7 to 8 years. And the message here is that there were no emerging safety concerns whatsoever. Next slide, please. And then here is efficacy. Now just to remind you again, this is just during the open-label extension phase. So at this point, all patients are on evolocumab, but you see in blue are the patients who were initially allocated to placebo and then red, those who were allocated to evolocumab. And you can see the earlier initiation of treatment with evolocumab then led to substantial reductions, not only in the primary endpoint, there was a 20% reduction in the key secondary endpoint of cardiovascular death, MI or stroke, and perhaps most importantly, a significant reduction, 23% reduction, in cardiovascular death now that we had sufficiently long follow-up. Next slide, please. So the data I showed you were comparisons from OLE based on the initial allocation to placebo or evolocumab. A related question though is what is the optimal achieved LDL cholesterol with regard to cardiovascular and safety outcomes. Basically, what information can we glean in terms of when we write guidelines, what should we target for our patients. And so we explored the relationship between achieved LDL cholesterol, now agnostic to treatment arm, and the occurrence of long-term adverse cardiovascular and safety outcomes in the pooled FOURIER and FOURIER-OLE datasets. Next slide, please. And so here, you can see for both the primary endpoint on the left and the key secondary endpoint on the right, a continuous, really monotonic, essentially linear relationship, between achieved LDL cholesterol and the risk of cardiovascular outcomes. The lower the achieved LDL cholesterol, the lower the risk. And you can see that extends down. So as you go from 70 to 60 to 50 to 40 to 30 to less than 20, the risk for cardiovascular events gets progressively lower. Importantly, there is no inflection point here, right? You can't get your LDL cholesterol too low in these individuals. Next slide. And now, conversely though, for safety outcomes, and here, we've been with patients in the 6 different groups in the legends shown at the bottom there, there is no relationship between achieved LDL cholesterol and the risk of any of these 8 safety outcomes that have been looked at, either in relation to statins or to low LDL cholesterol. We see no relationship whatsoever. So that's reassuring. Next slide. And so for this part, I'll summarize that long-term use of evolocumab with a median follow-up of more than 7 years appears both safe and well tolerated. The earlier initiation of evolocumab is associated with continued accrual of cardiovascular benefit, including importantly, cardiovascular mortality over the next several years. And we observed a monotonic relationship between lower achieved LDL cholesterol levels, down to very low levels, under 20 milligrams per deciliter, and a lower risk of cardiovascular events, up to 8.6 years of follow-up. These data are very timely because the most recent ACC Expert Consensus Decision Pathway on LDL cholesterol lowering now recognizes this principle. And in that document, they state there appears to be no LDL cholesterol level below which benefit ceases, again reaffirming the notion that you can't in these patients get your LDL cholesterol level low enough. Next slide, please. So now, I'll transition to data that we also presented here at the American Heart for reductions of lipoprotein(a) with a small interfering RNA, and these are the results for OCEAN(a)-DOSE TIMI 67. Next slide, please. So yes, thank you. So Lp(a), just to orient people, is a low-density lipoprotein, but it's a special one in which apolipoprotein(a), a specific protein, is bound to this LDL particle. On the left, I show a histogram of Lp(a) levels in a population with atherosclerotic cardiovascular disease, and you can see the long right tail for that distribution. And just by way of a benchmark of an Lp(a) at/or above 150 nanomoles per liter, that's in about 20% of the population. On the right, I show the relationship between Lp(a) levels and the adjusted risk of atherosclerotic cardiovascular disease adjusted for traditional risk factors. And you can see here a continuous relationship between those, and that the higher the Lp(a), the higher the risk for cardiovascular disease. Not shown here, but there's also data from genetic studies from Mendelian randomization studies also showing that relationship. The challenge is that up until recently, we didn't have therapies, dedicated therapies, that could robustly lower Lp(a) levels. So next slide, please. So that brings us to the study scheme for OCEAN(a)-dose. And in this study, we enrolled 281 patients between 18 and 80 years of age, with known atherosclerotic disease, and an Lp(a) level greater than 150 nanomoles per liter. Patients were divided into 1 of 5 arms for Olpasiran arms and 1 placebo arm. For the Olpasiran arms, they received either 10 or 75 or 225 milligrams every 12 weeks. And there was a fourth arm that also got 225 milligrams, but in this case, every 24 weeks rather than every 12 weeks. And the primary endpoint was the percent change in LP(a) from baseline to week 36. Next slide, please. This just shows the baseline characteristics. For the sake of simplicity, we've pooled the Olpasiran arms. This is a population typical for atherosclerotic disease. The age is in the mid-60s. About 1/3 were women. The majority had coronary disease. And roughly 10% had peripheral arterial disease, and about 20% also had cerebrovascular disease. You can see it was a well-treated population. Close to 90% were on statin therapy. About 1/2 were receiving ezetimibe, and about 1/4 were receiving a PCSK9 inhibitor. And then, by design, the Lp(a) level was high. And in this case, it was around 260 nanomoles per liter. And then reflecting how well the population was treated, the median LDL cholesterol was just under 70 milligrams per deciliter. Next slide, please. And so this shows the changes in Lp(a) through follow-up, and I'll walk you through this graph. The top line in gray is what happened in the placebo arm. And basically, the level stayed very similar, just a slight drift up about 3% to 4%. And then we can see, in blue, the Olpasiran 10-milligram every 12-week dose, where there was a roughly 60% to 70% reduction in Lp(a) levels. Then when we get to the 75- and 225-milligram doses every 12 weeks, now we're seeing reductions on the order of 90% to 95% and then greater than 95% for those 2 doses. And then in the yellow-orange color there, you'll see the Olpasiran 225 every 24 weeks, where you can see that then, by the end of that 24-week dosing interval, is a slight drift back up in Lp(a) levels. Next slide, please. And so for the primary endpoint, which is the placebo-adjusted change, meaning we took the percent change seen in each Olpasiran arm and subtracted from that the change seen in the placebo arm, you'll recall the placebo arm went up by a couple of percentage points, and so therefore, once one exceeded doses of 75 milligrams every 12 weeks, we achieved greater than 95% reduction in LP(a) levels. Next slide. And this slide here shows the waterfall plot. And now, you're seeing the effects on each individual patient in each of these arms. And what I'll underscore here is that for the doses of 75 milligrams or higher, there is a striking consistency in the effect, so really minimal interindividual variability for the effects of the small interfering RNA. Next slide. Now for safety and tolerability, you can see here, at the top, the rates of adverse events overall. There were no differences. There were no differences in the rates of serious adverse events. And I think that the metric that I find clinically most important as what were the rates of adverse events that led to discontinuation of study drug, and those were quite low, under 2% in all of the arms. And then for side effects that people have worried about, in general, for sort of lipid lowering in the field, myalgias, liver-related adverse events, hyperglycemia or new onset diabetes, no dose response throughout. There were higher rates of injection site reactions and hypersensitivity reactions, but these need to be qualified. So first, the injection site reactions were, by and large, mild, and the hypersensitivity is a function of MedDRA coding. It's not what clinicians, I think, would consider hypersensitivity. These are, by and large, injection site pain, but through how MedDRA worked, they sort of rolled up to that. But these were not like an allergic reaction, like a bee sting, that you might think of for hypersensitivity. And again, these all resolved without any treatment typically over 24 or 48 hours, so I think a very reassuring safety profile. Next slide, please. So in conclusion, Lp(a), we know, is associated with atherosclerotic cardiovascular disease risk, independent of traditional risk factors. Olpasiran is a small interfering RNA. And when dosed at 75 milligrams or higher every 12 weeks, really robustly reduces LP(a) levels by more than 95% in patients with established atherosclerotic cardiovascular disease, with really minimal interindividual variability for that, so quite consistent effects. Olpasiran appears both safe and well tolerated in this study, and these findings then set the foundation for a Phase III testing to begin later this year. Next slide. And so these results were just published in the New England Journal of Medicine as you can go there to read more about this study. And next slide, I think then this just sets the stage then for OCEAN(a) outcomes. As I alluded to, that will be the definitive outcomes trial with Olpasiran, where, again, we're enrolling patients with established ASCVD, with an elevated level of Lp(a). And they will be randomized Olpasiran or placebo and then followed up. The primary endpoint is time to first coronary heart disease death, MI or urgent revascularization. And we estimate the study will start next month. I think that's the end of my part. So I thank everyone for their attention. I'll turn it over to Narimon.

Thank you, Dr. Sabatine. I am pleased to share an update on one of Amgen's early pipeline programs, AMG 133, an investigational therapy being evaluated as a treatment for obesity and obesity-related illnesses. Next slide, please. Recall that AMG 133 is an antibody-peptide conjugate. It has an antibody backbone that blocks the GIP receptor, and bound to this backbone are 2 GLP-1 peptides that stimulate or agonize the GLP-1 receptor. So it's a bifunctional molecule. The molecule has been engineered in this way on the basis of: one, human genetic data that our colleagues at deCODE contributed to, which strongly suggests reduced GIP receptor activity, reduces body weight; and two, the existing body of knowledge and important clinical experience with GLP-1 agonists. Our preclinical data suggested synergistic effects with GIP receptor blockade and GLP-1 receptor agonism on weight loss in a number of other metabolic parameters that led us to advancing this molecule to Phase I clinical testing. Next slide, please. And as shared earlier in the year and shown here, the single-dose cohort data from Phase I has supported the concept that these 2 mechanisms, blocking 1 receptor and agonizing the other, can meaningfully impact weight reduction. Next slide. So we're pleased to share that we've made progress with the Phase I study and have now completed both the single- and multiple-dose cohorts. Within each of the multiple-dose cohorts, a patient received a total of 3 subcutaneous dosing administrations, each 4 weeks apart, with the last administration just before day 60. We continue to see encouraging data from the Phase I study, specifically with multiple dosing. We've seen mean percent changes in body weight, ranging from roughly 7% at the lower dose, to 14.5% at the higher dose by day 85. No plateau on weight reduction was observed during the treatment period. We have not observed notable safety concerns with the most treatment -- with most treatment-emergent adverse events having been mild and transient. The pharmacokinetic data were dose-proportional and demonstrated an extended half-life. And what I'll say about the data on the whole is that we're encouraged. We're encouraged by the weight reduction, the pace of the weight reduction and the durability of the effects seen in the Phase I. And we look forward to sharing these data in greater detail at the 20th WCIRDC Congress in December. In the meanwhile, we continue our plans to initiate a Phase II study, where the full attributes of this molecule can be elucidated, and we intend to start that early next year. Next slide, please. So moving on to our conclusions. To conclude, we pursued a strategy of prosecuting genetically validated targets in our cardiovascular and metabolic pipeline, ranging from PCSK9 to Lp(a) for atherosclerotic cardiovascular disease, to GIP receptor for obesity. And the results we've seen so far with this strategy are encouraging. The Repatha FOURIER open-label extension findings outlined by Dr. Sabatine today are compelling evidence of the impact of early and significant LDL cholesterol reduction and how this can impact cardiovascular events. The Olpasiran Phase II data show a rapid and sustained reduction in Lp(a), and we're looking forward to initiating the Phase III outcomes trial by the end of this year. And lastly, with AMG 133, a Phase I program, we've demonstrated significant dose-dependent reductions in weight of up to 14.5% by day 85, with an every 4-week dosing regimen. And we plan to initiate a Phase II dose-ranging study early next year to better understand the full potential of the molecule. This concludes this section of the presentation, and I believe we'll be turning it back to Dave. Dave?

Thank you, Narimon. And before that, thank you, Dr. Sabatine, for an excellent review. [ Jay Son ], do you want to go ahead and review the process for asking questions with our listeners? And then we'll go ahead and open it up to the Q&A session.

[Operator Instructions] Our first question is from Michael Yee with Jefferies.

Obviously, I guess, I'm going to go with the 133 question. David, thanks for the update there. Given the commentary, can you just maybe conclude where you see the strongest differentiation? Is it on efficacy? It appears that a negative 14% are dosing once monthly or the kinetics. Or maybe just sort of summarize that and what you think is most important there.

Yes, Mike. Thanks for the question. And of course, we knew there would be a lot of interest here. I think potential points of differentiation here are really what you've articulated, dosing interval, potential efficacy. Obviously, we need to further test that in Phase II, since the Phase I trial only included up to 3 doses. But we did not appear to see a plateau in dosing at the conclusion of 3 doses. And in fact, that weight loss was relatively sustained. We're continuing to follow most of those patients. And then, finally, the kinetics of weight loss, meaning the rapidity of weight loss, we think, is encouraging. And that's something that we will also be examining carefully in Phase II. So all of these are potentially important attributes, and we'll be generating more data in what will be a robust Phase II study.

Our next question comes from Matthew Harrison with Morgan Stanley.

I was wondering, if you could just comment maybe a little bit more broadly on Phase II plans here for 133. I guess 2 parts to this question. So the first one is do you think the only path to registration here is a large outcome study? Or are there certain populations where maybe there could be a faster path to market? And then, secondly, maybe you could just comment on the dose range that was chosen in Phase I and if there was any sort of upper bound related to potential safety concerns there. I think you can -- you may push the dose higher in Phase II.

Yes. Thanks, Matt. I'll ask Narimon to comment and provide a little bit more in terms of specifics on the Phase II design. What I will say is that one of the things that we've really taken into account for Phase II is preserving optionality. So it will be a relatively large study. It will include patients with diabetes or glucose intolerance and without that. And we'll look at a variety of doses and schedules to really then guide us for further development. I think it's probably too early to speculate on whether earlier registration could be achieved. That's something that we will, of course, take a look at. And as Phase II proceeds, we'll provide guidance on our thinking on what that pathway may look like. But let me turn it over to Narimon to provide a little color on our Phase II thinking.

Thanks, Dave. So we'll be pleased to share the details of our planned Phase II study once finalized and available on clinicaltrials.gov. But at a high level, it's planned to be a robust dose ranging study of several hundred obese patients with longer follow-up than what we had in Phase I. We'll be evaluating a range of doses as well as a variety of dosing schedules, including fixed doses with fixed interval dosing as well as some dose escalation. And we'll have both patients with and without diabetes included. And in terms of what we'll be looking at, we'll be looking at a course rate reduction as well as a number of other clinical parameters.

Great. Thanks, Narimon. And we will be sharing plans on that very soon. But again, we want to preserve optionality with this study, and we believe it's going to be quite a robust study design.

Our next question comes from Chris Schott with JPMorgan.

And just staying on the 133 questions. Can you just comment at all or elaborate on the HbA1c benefit you saw with the drug from the Phase I? And maybe just more broadly, can you talk about how you'd expect this agent to perform on HbA1c, I guess, relative to Mounjaro, given the high overlap in the obesity and diabetes populations?

Yes. Again, I'll ask Narimon to comment on just a minute here. The Phase I was not necessarily focused on patients with diabetes. So I think we need a larger experience, and that's one of the things that we will be taking a look at, specifically in cohorts in Phase II, as I mentioned. It's also worth reminding all of us that weight loss itself can be an important part of control of type 2 diabetes. And if one is able to achieve ultimately significant weight loss, such as 20% or more, that can restore, for example, peripheral insulin sensitivity and has a host of other metabolic effects. Narimon, do you want to comment a bit further on the hemoglobin A1c question regarding Phase I?

Absolutely. Thanks, Dave. So preclinically, as you would expect, given the construction of the molecule, we did see effects on issues like insulin sensitivity, glycemic control and others that you would expect to be favorable in human -- with human testing. But as you pointed out, our Phase I study excluded patients with an existing diabetes issue or other significant comorbidities. So the patients were euglycemic when they came in to the Phase I, and we wouldn't expect to see significant differences on that basis. However, in Phase II, we will be including patients with diabetes and metabolic syndrome. And we will expect to see those types of efficacy metrics best assessed there.

The next question comes from Salveen Richter with Goldman Sachs.

Could you speak more to the tolerability and adverse event profile? And then help us understand what you'll be presenting in December at the Congress.

Great. Thanks, Salveen. Yes, I'll let Narimon speak briefly to tolerability. We don't want to give away the entire dataset for next month to honor the embargo from that meeting. But Narimon can provide a little color there on that and the adverse event profile. And certainly, we'll tell you what you can expect to see next month here in Los Angeles. Narimon?

Thank you, Dave. So let me just begin by caveating, our Phase I experience is early phase experience. So the types of cohorts and patient numbers that we look at in those cohorts are rather small. The totality of the Phase I experience was around 70, 75 patients across many cohorts, but I wanted to start off with that caveat. In terms of what we will be presenting in December, as Dave outlined, you'll see more detailed data from that Phase I. But what we can share today with regards to safety is that there were no significant or notable safety findings observed. And in fact, most of the treatment-emergent adverse events that were observed in the Phase I study were mild and reversible. Now in terms of the types of adverse events, the qualities of those adverse events that have been observed, I would say that they are typical of what you would see with the incretin class, with the most prevalent ones being related to GI effects.

Our next question comes from Jay Olson with Oppenheimer.

I'm going to ask about Olpasiran. The New England Journal publication indicated that it's unclear how much of a reduction in Lp(a) concentration would be required to translate into a meaningful reduction in the risk of cardiovascular events. Can you just comment on what you see as the target for Lp(a) reduction and what role that may have played in your dose selection for Phase III?

Yes. Great. Thank you, Jay, and thank you for asking an Olpasiran question. I'm going to ask Dr. Sabatine to comment in a minute here. The one thing that I would point out, if you recall, that the average Lp(a) level was about 260 in the Phase IIb study, so a 90% reduction of that brings you to, say, roughly 25 nanomoles. That is where many folks with so-called normal range Lp(a) live and so you're driving the Lp(a) into that normal range. Now of course, we don't have the 6 or 7 decades of history with Lp(a) that we had with LDL cholesterol. But part of our thinking was really to drive that into the normal to low-normal range in as many patients as possible. Dr. Sabatine, would you like to comment on goals in terms of Lp(a) reduction?

Yes, sure. And I'm happy to add. I obviously agree with what you said. I think the exact nature of the relationship between Lp(a) and cardiovascular risk remains to be fully defined in some studies, that it looked like that once you got down to a certain level, that if you were, as David said, in that sort of low range, that the risk was low, the exact threshold can vary. In part, it's also trying to gauge how much lowering, if it's like LDL cholesterol, how much lowering would translate into what percent reduction. There have been a variety of genetic studies, and they have given a variety of estimates. I think the good news for Olpasiran is that it reduces Lp(a) so much, that the effect should be quite robust. But I think these first studies will actually give us the data like we have for statins and LDL cholesterol, for that exact relationship. Otherwise, it's a bunch of extrapolation from genetic studies. But we should, in the outcomes trial, be positioned given how potent this drug is, to see a robust reduction in clinical events.

Our next question comes from Yaron Werber with Cowen.

Yaron, are you with us?

Yes. Can you hear me? I can hear you.

Now we do.

So maybe a question for David and Narimon. Just on 133, is the data ITT that you're giving us? Or is it on treatment? And then, secondly, when you look at some of the other drugs, [ they're deeper ] sort of partial agonists and you're an antagonist. Can you give us a little bit of a sense, what are you seeing in the [ cogenetic ] data between those 2 in the natural history?

Yes. Maybe I'll start with a perspective on the second part there and then ask Narimon to talk about some of the specifics in the ITT population in Phase I. As Narimon noted, there have been 3 large GWAS studies that have examined the relationship between variants that affect signaling through the GIPR pathway. And in general, those suggest that variance that are associated with reduced signaling through that pathway are associated with lower BMI and lower weight. In fact, the major contributor to those studies was us, with our team in Iceland. And so -- and we've, in addition, done additional analyses that really convince us that antagonism of the receptor is the appropriate approach given this wealth of genetic data. So that's really what led us directly to the construction of this molecule. In addition, it's on an antibody backbone, which gives it a potential favorable pharmacokinetic profile and lead to the sort of dosing intervals that we've talked about. Narimon, do you want to speak a little bit more about Phase I in response to Yaron's question?

Thank you, Dave. So to be clear, the results that we are citing for effects are on-treatment results, Yaron. So these are patients that were taking on therapy. Most of the patients, over the course of the Phase I, continued on therapy successfully. And to perhaps put a double underscore on Dave's points, which are spot on, we have conviction in following human data. And the human genetic data have been quite convincing that it is reduced GIP receptor activity that is associated with reduced body weight, not more GIP receptor activity. And so we have designed this antibody to bind to and block the GIP receptor. And we believe that the effects that we are seeing are reflective of that. Now that being said, there's obviously more opportunity to do mechanistic data preclinically to understand what happens with agonism and antagonism. But what has been published in the literature, and you can find, is some papers that show that chronic agonism eventually leads to desensitization of the pathway and maybe functionally something similar to blockade. But that work needs to continue. Thank you.

Our next question comes from Umer Raffat with Evercore.

This is Mike DiFiore in for Umer. Just 2 questions from me, both of them on AMG 133. Number one, I was wondering if you could comment on how baseline body weight could affect drug exposure with this asset. We saw from the semaglutide data that the higher the baseline weight, the less drug exposure. And I was wondering of how adding the GIPR mechanism may play into this dynamic, if at all. My second question is that given that weight loss didn't plateau, I realize it's early days with this asset, but I was wondering if you could offer any commentary on how you're planning to strike the balance between convenience with less frequent dosing versus optimizing efficacy and perhaps by sticking with every 4 weeks.

Thanks, Mike. I'll ask Narimon, of course, to comment on both of these questions. One thing I would point out, to start, is that, obviously, exposure response analyses will be something that are important for us to take a look at in the Phase II study as we've got a larger group of patients. One of the things, again, that I think is important about this molecule is that it does have an antibody backbone, which gives that antibody-like pharmacokinetic behavior. But I'll ask Narimon to comment on the issue of baseline body weight and potential changes in exposure and then the convenience issue. Regarding the latter, we will be, as Narimon noted, testing a variety of approaches, including fixed doses as well as titration regimens in Phase II to really preserve optionality for Phase III designs. Narimon, do you want to add some color?

Thank you, Dave. So insofar as our Phase I study is concerned, the baseline BMIs were approximately 32 to 34. So everyone that came into the Phase I study needed to be an obese person. Insofar as our pharmacokinetic data are concerned within the patients that we have looked at and within those ranges of weight, we have not observed notable differences across those patients in terms of their exposure. Now what we can also offer in terms of our dosing schedule is we feel quite comfortable with an every 4-week dosing regimen at a fixed dosing regimen. So we don't think that we necessarily have to be pitting efficacy against dosing convenience with this particular asset. Thank you.

Great. Thanks, Narimon. And of course, this field is relatively young over time. There will be a look at maintenance dosing and a variety of approaches. But I think we've got a molecule that gives us quite a bit of flexibility right now.

Our next question comes from Mohit Bansal with Wells Fargo.

So again, one question on 133. So is there any reason to believe that an antibody approach versus a peptide approach could have some advantage on safety as well? And then the other part of the same question is that did you employ any kind of dose titration in this particular study?

Thanks, Mohit. I'll ask Narimon to comment on the second part. What we can report so far regarding safety is what Narimon has discussed. And we're quite pleased with what we've seen so far. Obviously, those are datasets that will be fleshed out in Phase II. Narimon, do you want to add any comments here?

Thank you, Dave. So I'll just say all of the data that we have shared today have come from multiple dose patients that were dosed in fixed dosing intervals with a fixed dose. So there has been no -- none of those results have been reflective of patients that haven't gone through dose escalation or a dose titration, and that's inclusive of the safety results that we've also commented on. We continue to apply our knowledge to testing a variety of different dosing schemes, including potential dose escalation schemes, which we'll also be looking at in Phase II. So we're keeping our options open, but we feel quite comfortable with our fixed dosing intervals at this point.

Our next question is from Evan Seigerman with BMO Capital Markets.

This is Conor on for Evan. Just one question on Olpasiran. Can you talk about your expectations for the cardiovascular outcomes data in the Phase III trial given the impressive Lp(a) reduction you've seen in Phase II? And then on those expectations, what might you need to see in the Phase III trial to drive real-world uptake?

Yes, I'll ask Dr. Sabatine to comment on what he'd like to see from -- in Phase III. But this is an outcomes trial that's -- I think as it rolls out, you'll see that it's consistent with other outcomes trials in the field, and it's designed to show a clinically relevant reduction in major cardiovascular events. Dr. Sabatine, let me ask you to comment on what you'd like to see here and what you think clinicians will need to see in order to use this sort of therapeutic.

Yes. No, I think it's a great question. I would say, again, because Olpasiran is so potent at lowering Lp(a), that given the population we're studying, based on the data we've seen in the genetic studies, I think there should be, I'll say, a robust reduction in clinical events. That would be one that, as a clinician, if I saw, I would then want to give this to my patients. So the second part, I think then, is in terms of the uptake, and I think Lp(a) is being increasingly recognized as an important risk factor. The good news for it is that the levels tend to be fairly stable over time. And so with more and more professional societies recognizing this, and now, with drugs coming down the pipe that can then make a meaningful impact on LP(a) levels, I suspect more and more clinicians will start to be measuring Lp(a). And then obviously, once the results come out, and if it goes along as I anticipate, there will be a robust reduction, then I think clinicians will very readily embrace checking this for their patients.

Our next question comes from Robyn Karnauskas with Truist Securities.

I guess I'll throw a curveball and ask a Repatha question. So on the FOURIER long-term extension study, maybe Dr. Sabatine can talk to this as well, Amgen has guided to much more for Repatha. Can you talk to us about how doctors use extension studies? Do you think that more patients -- what will it take for more patients to get treated with the drug, both at the specialist level as well as the primary care level?

Right. Thanks, Robyn. And of course, I'll ask Professor Sabatine to comment in just one moment. To me, one of the lessons, the clinical lessons from the extension study is that not only is lower is better, but earlier is better, in that those who started Repatha later never quite caught up in terms of event rates. And to me, there's a lesson for the field there. But of course, let's have Dr. Sabatine comment on that and how you put this into clinical context.

Yes, happy to do so. It's a great question. I think the open-label extension data have been very well-received by clinicians for multiple reasons. I think, one, when we initiate lipid-lowering therapy, we envision then it will be lifelong for the patient. And so we were spoiled in the statin era for having trials that, as I noted, had a median follow-up of roughly half a decade. But we find that duration of follow-up comfortable or comforting to us when we prescribe a drug long-term. So OLE then gives us that. And in fact, as I indicated, for some individuals, even out to 8 years. So I think that's very reassuring and gives a long-term profile that now is at/or even longer than we've seen for statins. And then, secondly, I think for the efficacy, as I noted, that it takes time to see the full benefit for LDL cholesterol lowering. And I think OLE nicely illustrated that, with more time, you see not only a reduction in MI and stroke, which we had showed earlier, but now, reduction in cardiovascular death. And that's the sort of benefit that clinicians really sort of sit up and pay attention to. I think that's coupled with the most recent consensus statement, I mean, both the European guidelines, which embrace a lower target than we had in the U.S. as they target under 55 milligrams per deciliter for high-risk patients and even for those with recurrent events, to be more aggressive and be under 40. And then the most recent U.S. guidelines, again, embrace under 55. Note that there's no lower limit. And so as there's increasing recognition for that, and we want to for our patients, at least for secondary prevention, get their LDL cholesterol levels down, and we want to do that long-term and as early as possible, as David indicated, then in those cases, you're going to need a PCSK9 inhibitor to get there for your patients.

Our next question comes from Geoff Meacham with Bank of America.

I have a question for you on 133. So when you look at tirzepatide, it had benefits beyond just weight loss and A1c. It was looking at lipid lowering and blood pressure lowering. I'm just curious, if you have any data for 133 on those type of biomarkers. And then what you would be looking for on beyond just weight loss in the Phase II.

Yes. Thanks, Geoff. I'll take that question. So obviously, we expect the effects on a wide variety of potential metabolic parameters, and we're going to thoroughly study that in the Phase II. Based on what we've seen so far, it's encouraging. Remember that this is relatively short duration of therapy in Phase I, and so we need longer-term data from Phase II. But I think we're quite encouraged by what we've seen so far, and we will very thoroughly profile patients from a metabolic standpoint in the Phase II study.

Our next question comes from David Risinger with SVB Securities.

So could you discuss to what degree as an antibody or having an antibody backbone, as you described, that AMG 133 crosses the blood-brain barrier and impact [ society ]?

Yes. Great. Thanks, David. I'll take that question. Because of this molecule, it's a large molecule built on an antibody backbone, as we've said, we would not expect it to cross the blood-brain barrier or achieve appreciable concentrations in the CNS. So most of the -- or most, if not all of the effects that have been observed, we would expect to be due to peripheral activity of the molecule.

Our next question is from Colin Bristow with UBS.

So a couple more. On 133, were these figures on a background of diet and exercise? And what was the specific time point that the 14.5% weight loss was measured at? And there's one on Olpasiran. I see in Phase III, you've just -- you've moved the inclusion criterion up to 200 nanomoles per liter for Lp(a), I think, versus 150 in the Phase II. Just curious why you made that change.

Yes. Good questions, Colin. And I'll ask Dr. Sabatine in just a moment to comment on the second one on Olpasiran, why we chose the 200-nanomole cut point. And then I'll ask Narimon to comment on the diet and exercise question and the time frame. But recall that this is, as Narimon noted, the weight loss of up to 14.5% was by day 85 or so. Narimon, do you want to add any additional detail there?

No. Thank you, Dave. You got it spot on. It was by day 85 in terms of the effect. And just very quickly, with regards to diet and exercise, there was no specific instruction given with regards to diet and exercise in this Phase I study.

Great. Thanks. And Dr. Sabatine, would you like to comment again on the Olpasiran or Lp(a) threshold level that we've chosen?

Yes, sure. No, it's a great question. I would say it's analogous to what was done with statins, right, if you'll recall, for 4S, looking at patients with existing coronary disease and started with very high LDL cholesterol levels. And so it was 188 milligrams per deciliter. But that helped them define the magnitude of benefit. And so I think in an analogous way here, we want to start with the group with a very high level, that I think we would have a high degree of confidence that we'll see a very robust reduction. I think it's going to be quite robust. And then if it looks as good as I hope it will look, that would then open the door to testing populations with lower levels of LP(a). But you have to start at some place. And so we're starting those with high levels.

Our next question comes from Dane Leone with Raymond James.

Apologies for adding on to the most analyzed Phase I/II slide dataset that we've had for a large-cap biopharma company. Can you just clarify, just getting investor questions in, can you just clarify in terms of how you actually cut off the high and low dose within that chart that you have for the mean change for baseline in weight? Is that -- was there any cut point of what the doses were for high or low and the number of patients and each that kind of gives you confidence that there's real separation and not allow statistical noise in there?

Yes, sure. I'll ask Narimon to comment on that, in what I think you're asking is, in part, are we confident of a dose response. I would add, before Narimon speaks, that this is probably the most discussed, not yet presented, Phase I trial that we've had. So Narimon, do you want to comment on the dose response question?

Yes, sure. So Phase I data, typically, we're looking at [ precious few ] per cohort. In this case, there were 8, and we had multiple cohorts over the course of the Phase I. The dosing we'll disclose at the presentation, but every cohort had a specific dose assigned to it. So when we speak to low dose versus high dose, it speaks to the cohort that was at the low dose and the cohort that was at the high dose for comparison.

And you'll see all of that in some detail, Dane, in about a month. All right. We've got, I think, a couple of more questions in the queue. We'll go a few minutes over here if we have to, since we had a brief delay earlier.

Our next question comes from Matthew Phipps with William Blair.

You've mentioned that human genetic data was behind using a GIPR antagonist versus -- to go with the blocking antibody there. I was wondering how you felt about the potential impact on bone turnover given GIPR expression on osteoblast and osteoclast. And some GIPR variants with the [ mutations ] have been reported as they have lower bone marrow density?

Yes. Thanks, Matthew. That's a question that we've looked at directly. And I'll ask Narimon to comment on that. But I would say, right now, we're quite reassured based on what we know. Narimon?

Thank you, Dave. So we're familiar with this work. We've taken note of these observations. And our colleagues at deCODE genetics that do these types of analysis very rigorously have looked at this type of question, and they have not found -- they have looked and have not found such an association with fracture risks or bone mineral density reductions. So we feel quite confident about that observation. But we are also going to make sure that we're going to continue to monitor for things like bone mineral density. We have done that in Phase I. We have not seen notable differences to date, but we'll continue the monitoring into Phase II just in case we're missing anything. Thank you.

Our last question comes from Carter Gould with Barclays.

I guess just going back to the dual mechanism with 133. When you think about, we've seen with other related mechanisms, some sort of bounce back in terms of weight gain upon stopping therapy, can you maybe just frame how you think the dual mechanism here might compare on a relative basis? And I don't know if there's anything from your preclinical work or the patients who stop therapy in the Phase I that you could point to on that front.

Yes, maybe I'll handle that, Carter. As Narimon mentioned, we saw good sustainability. The other thing to keep in mind because this is built on an antibody backbone, you don't have a sort of rapid on-off with these pathways, and they may -- that may help, in theory, prevent some of the rebound that you've been discussing. Obviously, this is something we'll take a look at in Phase II, but we've been quite encouraged regarding that question based on the data that we've seen to date. All right. Well, I think that gets through all of the questions. I really want to thank all of you for joining us today. As always, the Investor Relations team will be standing by if you have additional questions. And I'm sure we'll see many of you here in Los Angeles for the presentation of the AMG 133 Phase I data in about a month and then at meetings throughout the remainder of this year and on into the first part of next year. So thanks, everyone, and have a good evening.

That concludes the conference call. Thank you for your participation. You may now disconnect your lines.