Good morning, everyone. I hope you're doing well. Really pleased to see a packed room and really that goes for right now and also just the entire conference. My name is Akash Tewari. I head our pharma and biotech efforts here at Jefferies on the research side, and we have Amgen on the panel today, so really excited to have a discussion. I'll hand it off maybe to Peter for opening remarks, and we'll get going.
Great. Thank you, Akash. Good morning, everyone. Great to see you. We are glad to be here. I'm not used to holding the mic. That's all right. We're really glad to be here with you. We appreciate your interest in Amgen, and a couple of thoughts for you here before we get going in the Q&A. We believe there are lots of ways to win for Amgen, especially for patients. We're pleased with the strong first quarter performance, which reinforces 2026 as a springboard year for Amgen. This is a year which we expect our rapidly growing products to offset the impact of the loss of expirations on our denosumab franchise and the increased competition there that will occur throughout this year. In the first quarter, we did exactly that.
Revenue was up 6% year-over-year, and non-GAAP earnings per share was up 5% year-over-year, demonstrating the disciplined financial management that is historic for Amgen. 16 products delivering double-digit or better sales growth, 17 products annualizing at $1 billion or more based on those first quarter product sales. As we've been saying for quite a long time, breadth and depth across our portfolio.
Importantly, our six key growth drivers, Repatha, EVENITY, TEZSPIRE, the innovative oncology portfolio, the rare disease portfolio, and the biosimilars portfolio, made up 70% of the product sales in the first quarter, and that group grew 24% year-over-year. We expect those to continue to drive us through the end of the decade. I would suggest in that inside the innovative oncology portfolio, we have IMDELLTRA now annualizing at over a billion dollars, a medicine for treating small cell lung cancer, and I'm sure we'll talk about that a little bit.
Sure.
Inside of the rare disease portfolio, we have UPLIZNA, which is treating NMOSD, neuromyelitis optica spectrum disorder. It's good for the CFO, isn't it?
You nailed it.
Not bad. Also IgG4-RD, which it was approved to treat, I believe last May, and then a year ago May, and then also gMG, generalized myasthenia gravis, approved in December, I believe. We're very excited about UPLIZNA. The growth drivers are doing very, very well. Turning to the pipeline, confidence continues to build in MariTide as a new paradigm for the treatment of obesity, type 2 diabetes, and obesity-related conditions. We're executing effectively across the enterprise, in the pipeline from clinical development to manufacturing.
As part of that, we're advancing MariTide's broad phase III program and build and optimizing our manufacturing capacity ahead of that launch. I've had an opportunity to visit all of the sites a number of times, and our great manufacturing capacity, our world-class manufacturing capacity is coming along really well. We expect MariTide to be really important for weight loss induction, long-term weight maintenance, patients switching from weekly GLP-1 therapies to MariTide with the potential for as few as four to six injections per year. Our approach reflects how obesity care may evolve. Patients need effective initial weight loss, but they need practical, durable maintenance options.
Beyond MariTide, we have a robust phase III pipeline, including olpasiran, which is being investigated for the reduction of CV risk in patients with elevated Lp(a). I would note Narimon is a cardiologist in charge of our development, so I'm sure we'll want to visit olpasiran here, as important as that will be. Dazodalibep being investigated for Sjögren's disease, xaluritamig being investigated for late-stage prostate cancer, another bispecific T-cell engager. We're also investigating xaluritamig in earlier stages of prostate cancer.
In addition, we're pursuing new indications for several of our approved products, including UPLIZNA, which I mentioned previously, and autoimmune hepatitis and chronic inflammatory demyelinating polyneuropathy or CIDP, TEZSPIRE and COPD, and eosinophilic esophagitis and IMDELLTRA, again in early-stage small cell lung cancer. Now, Akash, I know you have a KOL or key opinion leader call, I think scheduled tomorrow on the tax-
Oh, on the tax case. Yeah, that's right.
I just wanted to mention on the tax front, since there have been some questions around that since our first quarter call. Two related matters with the same core issue, value attribution. The IRS is arguing that our Puerto Rico operation should be treated as if it were essentially a limited value contract manufacturer. We strongly disagree with that position, and we have since they've adopted it. We've got 2,500 plus colleagues who are U.S. citizens at our flagship facility there in Puerto Rico, the majority with technical degrees.
It's a very complex operations. Remember, there are two tax periods in question now, 2010-2015 and then 2016-2018. 2010-2015, that tax court matter is not new. It's been around for a number of years. There have been no changes in our position on the status of that case. It's been fully tried and litigated. The tax court stopped hearing that case in January 2025. We don't expect a decision from the judge any earlier than the second half of 2026. As fully expected in the ordinary course of the IRS audit process, we received a notice of proposed adjustment, or a NOPA, as I call it, for similar transfer pricing issues for that 2016-2018 period.
It's just an early step in what we would expect to be a multi-year process. It's not a final determination. Importantly, our commitment to Puerto Rico continues to grow. Since our first quarter call, we announced, you may have seen it, an additional $300 million investment into Puerto Rico on top of the $650 million that we've announced also in the past year or so. Puerto Rico is a core part of our operations, our net manufacturing network flagship facility, advanced manufacturing capabilities for biologics, deep expertise all the way around.
We disagree strongly with any characterization of those as our operations there as limited value contract manufacturing. We continue to believe the IRS claims are without merit and our reserves are appropriate as we have all along. Just closing, we're executing against the plan we laid out for 2026. The first quarter demonstrated the strength, the breadth, the depth of the business. Six key growth drivers are performing, phase III pipeline moving along, and we're maintaining our rigorous financial discipline on behalf of our shareholders as we're investing for the long term.
Understood.
Lots of ways to win, Akash. I just wanted to share those comments.
No, I appreciate that. That's excellent. The last panel I just hosted was an AI panel. I'll give a brief question on this, Peter, to you. I think we often hear about AI for drug discovery, but we had Albert from Pfizer. He's saying, "Look, right now, for large organizations, it's about how we operate our cost bases much more efficiently, and that's going to be the first wave of change." Really, here's the question China is making drug discovery cheaper.
The cost of making drugs has gone down. Commercial organizations are becoming more efficient. Yet, when I look at my model, I'll criticize myself or really the street widely, we are not modeling these businesses, including yours, being a step order more efficient when we think about the margin structure, right? We still have the 20%-25% on R&D. We still have the 20%-25% on SG&A, and then we have the corporate costs, which can often be very substantial. Are we making a fundamental mistake here, and we are going to see enormous improvements in operating structure with companies like yours in the not too distant future?
Thank you. It's a very important question. We are fully committed to technology, artificial intelligence, data. I would suggest we're fully committed to it across the enterprise, and I will invite Narimon in a moment to discuss ways in which we are using it and applying it in our research and our development functions. I have said all along, Akash, that in research and development, becoming very, very proficient, if you're one of the leading biopharmaceutical companies, which we are, we better be good at it.
We better be on it and working it as effectively, efficiently as we can. As you go across the enterprise, if you would allow me before I turn over to Narimon, manufacturing, we opened Amgen Ohio, which is a finished drug plant, a couple of years ago, finished drug processing plant, and that is our most technologically advanced plant. Roughly speaking, we have maybe 50% or 60% of the headcount we would have had in the plant before that.
By the way, the labor's fabulous in Ohio. We're super pleased with it. We're actually opening another plant there. It's under construction. AOH2, right next door. We called it a butterfly plant, just open the wing up. The technology, the AI there is fantastic. We have, it's either seven or eight automated ground vehicles. Nobody drives forklifts anymore t hey're automated, n o injuries, t hey don't go on break, t hey haven't been getting sick, t hey get recharged, and plug themselves in when they've got time to do it. We move over to the commercial operations, and we're using it very effectively there. Kave, I'd invite Kave after Narimon speaks, just talk briefly about it. I know we're taking some time, but this is important.
Yeah.
Akash, I would suggest, I've guided to, on behalf of Amgen, a 45%-46% operating margin this year. Look, as part of that, we have to start, and are incorporating, some AI to become more efficient. I would suggest AI is going from the individual and functional use cases across companies, including ours, to enterprise use cases. We're after that w e are pursuing it vigorously w e think it's really important w e realize too that it's going to cost more. When Google does an $80 billion equity offering, somebody's going to be paying for the cost of all this compute at some point.
Sure.
We're all going to be thoughtful about that. We feel it's incumbent upon us to get those returns for our shareholders, and most importantly to use it. Innovative medicines to patients better, cheaper, and faster.
Understood.
With that, Narimon, maybe a sentence or two. I know we're taking some time, but this is so important.
Yeah, no, thank you, Peter, and thank you, Akash, for the question. As Peter mentioned a moment ago, we're using artificial intelligence, starting with the very beginning of the R and all the way to the end of the D in R&D. This gets into helping us develop new insights, improving our speed, cost, and quality of the work that we do in R&D, all of which are very important. To give you a sense of some examples of that, in terms of insights, think about the data that we have had through our work at Amgen deCODE and looking at multi-omic data.
We're looking at human genomic data, proteomic data, pulling all that information together and coming up with new targets to interdict upon developing new medications. That is now going to be accelerated and assisted with AI, new insights. We also use this technology to help speed up our late antibody optimization efforts in developing the therapeutics, right? We've seen improvements in speed of up to 50% on that angle in developing the therapeutics to then take into the clinic for clinical testing.
Once we make our way into clinical studies, we've applied artificial intelligence in helping us select the sites tailored to the population of patients that we're recruiting, and at times we've seen a threefold increase in the speed of recruitment on the basis of picking the right sites that are tailored, again, to the right drug for the right trial.
Once you're finished with the clinical studies, there's of course another R, and that is regulatory and filing, and we found that artificial intelligence has enormous potential in ingesting all of this data that we've generated over the life cycle of a medicine and putting it in the form of a draft document that we'll be ready to submit to a regulatory authority for review. You can see through all of those steps, you get new insights i t helps with the speed, obviously. The quality of the work that we do goes up, and that will ultimately help us reduce cost. Kave?
Yeah. From the commercial standpoint, to build on what Narimon and Peter shared, we're really focused on the use of the technology to speed up the access and use of our medicines around the world. We've got three big areas where we're focused on AI right now. First, remove the friction from getting access to our medicines both in the U.S. and abroad, going along with the regulatory comment that Narimon made i t's great if we can get the medicine approved faster.
We also need to get it paid for faster, using the AI to speed up the filing of our reimbursement applications, the quality of those reimbursement applications, in the U.S., access to medicine in terms of understanding where you are in the insurance scheme, co-pay, prior authorization, patient support. That's one big pillar, and that's going to show up in the revenue side more than the cost side overall. Second area is in patient identification, which is really important in rare disease. We've built some really great capabilities of identifying where patients are so that we can interact with the physician around the time that patient is seen.
We're using the AI to speed up and enhance those capabilities, which allows the effectiveness of our promotion to go even higher in those areas. Third, we know that physicians are using AI to change their own behavior. Making sure that our content and messages are showing up in those workflows in the AI of their choice is helping us identify and get patients onto therapy faster. Akash, when you think from the commercial perspective, we're going after the efficiencies, but those efficiencies are being reinvested to drive the top line faster, more than the cost line down at a differential rate.
I think that's an important point to note, and I think you'll hear that pretty consistently. It's more about top-line growth rather than just outright cost cuts. Maybe going into the clinical side, and we'll start with Lp (a). To me, you look at the genetic data, it's clearly a predictive marker LDL-C is such a rare biomarker in atherosclerotic cardiovascular disease because seemingly the more that it goes down, it continues to have a benefit. I always worry for Lp(a) there might be more of a threshold dynamic, where you get to a certain range and you can get into a range where it's not really impacting your cardiovascular risk.
There's more threshold dynamic rather than a marginal dynamic. We've seen headlines from Novartis and others in the field that these trials all tend to go longer and longer. I'd love to get your take about these trials going longer than expected, the difficulties in terms of modeling event rates in a modern cardiovascular population, and should we interpret the fact that these studies are going longer negatively or not?
Yeah. Thank you for the question. It's a really good question. Let me start by reminding us that we've had over 50-60 years of the benefit of looking at LDL, understanding LDL biology and the science, and thank goodness we're finally getting to the point and concept that lower is better, lowest is best. Right?
With Repatha, I think we have developed a very compelling evidence base, a mountain of evidence if you will, that shows that that is in fact the case. One of the important lessons from that half century of understanding LDL science and LDL biology is that it is the totality of LDL cholesterol burden reduction that is important for patients. Think of it this way i f you have a patient who has had a very high LDL cholesterol, carrying a large load for a very long period of time, that person is at high risk for having atherosclerotic cardiovascular disease, a heart attack. Right? What you want to do for that patient, and the science has borne this out and reduced it to a very simple equation from the CTT data.
is you want to remove as much aggregate LDL cholesterol burden from that patient for as long as possible. Right? If you look at the math, the equation is for every 38 mg/dL reduced LDL cholesterol, you have a 22% reduction in relative risk for major adverse cardiovascular events. That's the reductionist view of LDL cholesterol, and it's what makes a PCSK9 therapy like Repatha so powerful.
Right? Start earlier, take it longer, reduce the LDL to as low as possible. That's where the puck is headed for LDL biology. How do we apply that to Lp(a)? Your question. All right w ell, Lp(a) and LDL have a lot in common i f you look at the particle structure, they look quite similar, for the exception of a covalently bounded, more atherogenic and thrombotic protein LDL of Lp(a). Okay?
The way that we've approached studying Lp(a) is we want to start with the patients that have the highest burden of disease, highest burden of Lp(a), that are at risk and have shown that they have had untoward cardiovascular events, specifically coronary events, and we want to be able to follow those patients for a longer period of time. That is exactly how our study has been designed.
High Lp(a) threshold, more than 200 we have an incredibly effective medicine with olpasiran that can reduce that Lp(a) by more than 95%. In some cohorts from our phase II we reduced it by 100%. You want to have this type of therapy for those patients that carry such an enormous Lp(a) burden for a long period of time and have already suffered the consequences of poorly controlled cardiovascular risk. That is the basis of the outcome study that we have designed. As you said, Akash, the human genetic data, Mother Nature's experiment, the human epidemiology, all the anecdotes that we're seeing in the field, all point towards this being a very important dimension of residual cardiovascular risk.
Once you address LDL cholesterol, you absolutely have to address your Lp(a), and 20% of us in this room, unfortunately, have a high Lp(a), and I really hope that all of us have at least taken efforts to measure it because we have taken so long to understand what LDL values are for ourselves. That is simply not acceptable in today's age with the technology that we have. What do we make of the slower event rates? Should we be discouraged? Absolutely not. In fact, what we've seen over time.
If you go back to the 1980s, to the 1990s, to the 2000s, the event rates for accruing cardiovascular events have slowly gone down in the population of clinical studies. Why is that? Well we demand them to be on better medicines, for one. You want to see patients on PCSK9 inhibitors in a cardiovascular outcome study so you can understand what value your therapy brings forward on top of LDL reduction. That's one reason for it t here are a number of other reasons that could include a highly efficacious mechanism with Lp reduction. Because when you're looking at event rates in a study
you don't know which arm's having the event rate reduced. You're looking at an aggregate value. If you have a very effective therapeutic, the aggregate value of the event rate accrual is going to be lower. That is just going to be the simple math out of that. I don't really take any negative signs or concern- out of the event rate that's coming out of these studies. I think it's a natural outcome that we expect to see from better care and potentially very effective medications being brought into the clinical trial arena. I remain very enthusiastic about the mechanism, and I'm looking forward to see what the Novartis study will tell us.
It's a very clear answer, and I really do appreciate it. Hitting on MariTide, and really, I think there's an interesting scientific question because your team has talked about an antibody is different than a peptide in terms of how you can desensitize a receptor. I think Jay talked about it marinating the receptor in an interesting way. I wanted to think about that phenomenon versus you look at Novo, you look at Lilly.
These are companies that are more sophisticated than anyone in terms of running these trials because they have the experience. I remember going to ADA two years ago when the amycretin data came out, and it's like 50% vomiting on all of their amycretin arms. They're moving forward to phase III. Even with orforglipron, the issue of geographic variability came up pretty predominantly because your diet can really impact your nausea rates.
When I look at your drug and MariTide, really provocative data that you showed with a different kind of titration regimen. That was 140 patients in sites in Texas w e have to extrapolate that phenomenon to a worldwide perspective, and that's really the question a lot of investors have is when you think about geographic variability and just the complexities of running these obesity studies, which is seemingly getting harder, not easier because the adoption's gone up. How do you think about that phenomenon relative to also that really provocative data you've shown that perhaps an antibody has a different type of desensitization that maybe a peptide wouldn't? How should we think about that?
Thank you. Again, a great and insightful question on your part, Akash. Let's start with some basics. There are over 1 billion people in the world that suffer from this disease that we call obesity w ell over 1 billion. That number is not coming down. We look at the tremendous success we've seen from the newest therapeutics entering the field, and we see that the usage is about 2% or less of the addressable population.
There are a lot of people that have yet to even be touched by a medicine that can treat their obesity, and they really need this medicine. As we think about the whole space of opportunity, there is still almost all of the opportunity that still resides out in the market for patients that need a treatment for obesity. The opportunity is remaining very large. It's true that the scientific bar has gone up, has raised from five years ago. The opportunity remains very large, and there absolutely are going to be segments of the population, which I think you're getting to, that are going to benefit more or choose one therapy over the other.
MariTide has a very competitive and compelling profile from all the data that we have accrued thus far. phase I, phase II, and our very encouraging experience in phase III. We have what we believe is a very compelling and competitive efficacy as well as tolerability profile. More and more people need to understand that these medicines for chronic long-term conditions need to be taken for a long period of time. Right?
You benefit from medicines, including medicines that reduce LDL cholesterol, by taking them for a long period of time. It reduces your burden of disease over your lifetime t hat's how you appreciate the benefit, not just on weight, but all the comorbidities that are chronic comorbidities, like heart disease, liver disease, diabetes, from them. Why is this important? It's important because you have to think about what are the attributes of your medicine and intervention that enable a patient to take it for a longer period of time.
When you talk about a medicine going from being administered daily, which was back in the day of the first GLP-1s daily, to then weekly, that was a big innovation. That was helpful i t's going to be very helpful for people to go from weekly to monthly, potentially every two months, and potentially every three months, right? These are the questions that we are interrogating with MariTide. We start with monthly. We've recently announced two studies that are going to look at maintenance beyond the initial phase of treatment that will look at every eight week and every quarter administration.
You mentioned the switch study, where we're taking patients that are starting on their weekly medications, and again, we're going to help them understand, and the world understand, what happens for patients when they take those weeklies and decide, y ou know what I want to go to every two month or every three month administration. This will help with a large segment of the population, irrespective of the geography in which they reside. These are compelling basic needs of patients around the globe. Obesity, like many other diseases that we may end up talking about over the course of today, is a condition where patients will need more options, not less options.
By the way, I will interpret your answer this way, and I think it's quite important. The sense I get, and I think it's a fair point, which is we're so obsessed with these kind of contrived 68-week studies with titration profiles that no one actually adopts in the real world. Your point is, when you look at maintenance data, when you look at switch data, when patients are already on a background GLP-1, there's tons of data that can show the drug is very tolerable to switch. That, in and of itself, is an opportunity. Sounds like we don't know the answer to this question we'll get in the clinical trial, but we're thinking about this the wrong way is maybe the right way to.
Absolutely.
Yeah.
You're absolutely right. Think about this. Half the people today, of those 2% that have access to these medicines, stop taking the medicine before a year is up.
Yeah.
That's not great for those patients. We absolutely have to address the need of those people.
Understood. Kave, this is a transition to maybe when you think about launching MariTide and the commercial opportunity. It's interesting. There's just seemingly this obsession with this race to the bottom on price. I always feel like, whether you're Novo, Lilly, or Amgen, you're thinking about a portfolio approach. You're thinking about sequencing patients on multiple therapies. Also, the entry point might be $150 at a low dose. By the time MariTide is getting on the market, you have potentially millions of patients who are GLP-1 experienced who are now in the insurance market, right? One of the things that I think stands out about your development program is you're running the outcome studies. You're getting clinical data as well.
When you think about MariTide in the U.S., should we think about it not we're thinking about coming in at direct-to-consumer $150 to compete with oral Wegovy, but no, it's going to be patients who have already failed a GLP-1, often an oral or Wegovy, and now they are going to be in the traditional insurance system. Is that the right approach to think about in the United States? When you think about ex-U.S., launching an injectable medication, but it's a huge market, how do you launch an injectable product well in the rest of the world market, forget even just EU5?
Yeah. You covered a lot of ground there, Akash. Let me just say, I think that first, having a great partner like Narimon to really work through these issues in advance in development really helps us from a launch perspective. We think we've got the data that's going to, over time, show that it's easy to start on MariTide , it's easy to stay on MariTide, it's easy to sustain that weight loss on MariTide.
We're going to be ready for a number of different kind of segments of that market, whether that's a patient-directed market, an insured market, to your point, throughout. I'd say one of Amgen's strengths has always been the tight intersection with access through our development plans and our commercialization plans. Understanding the data and the economic drivers of a payer versus a patient is something that we are building that evidence base of. Depending on how the market evolves over the next few years, and it will evolve quite a bit.
we're ready for that market. I would say thus far, what we've seen from a pricing standpoint is in line with our expectations of what we thought would happen, and we are seeing in this market, which is different than a lot of pharmaceutical markets, that the price reductions are coming with demand lifts overall, and that's directed primarily by that patient-driven segment. Irrespective of where the patient is, we want to have a solution for them.
Yeah.
The development plan is helping us do that for the U.S., but also for outside the U.S., where some markets will be more patient driven and some will be more reimbursement driven.
Just hit on that. In line with your expectations, actually, that isn't totally dissimilar to, I think, how Lilly would describe it. There is this sense, I think, with the established players now that we are maybe out until the end of the decade to, at least in the United States, a level of pricing stability. We'll have mid-single digit declines. They seem quite confident that price of a cup of coffee, you've kind of reached that price point, that there is a ton of inherent demand just to have. Really the work's going to be on insurance access and just negotiating those contracts. Is that maybe the view Amgen has? Again, MariTide is, you're going to be thinking about that commercial market pretty shortly.
Yeah. I think that's a scenario that's out there, Akash. We're ready for a lot of different scenarios.
We're really looking to see how they play out in the market in advance of our launch. I think we're going to be ready for those scenarios once we get to market.
Understood. I'm going to sneak in one more question on IMDELLTRA. It's really about the opportunity in maintenance. It's funny, a lot of people focus on KEYTRUDA in lung, but I think a lot of the money they make is in a maintenance setting. I don't think that's well appreciated. Would it surprise investors if the maintenance opportunity in IMDELLTRA could actually be bigger than the frontline traditional opportunity there?
Yeah. I'm just back from ASCO. I can tell you that the physician community is extremely excited about IMDELLTRA in both the frontline and in the maintenance setting overall, and we're excited to be releasing additional data over time that shows that durability of response getting to long-term survival overall, Akash. I would say that data will still mature to inform the field, but I can tell you my interactions, the field's quite excited about the maintenance use of IMDELLTRA.
Understood. Thank you so much. I really do appreciate it. It's a wonderful discussion.
Thank you, Akash.
Thank you.
Thank you, Jefferies.
Thank you.
Thank you, all.