Amgen Inc. (AMGN) Earnings Call Transcript & Summary

Fantastic. Listen, thank you guys for joining us. I think it's become a bit of a custom now hosting Amgen Management on our conference. Pleasure to have the whole team. We have David from the R&D side. We have Peter from the finance office. We have Arvind from the Investor Relations and the finance office. So really great to see everyone. Arvind, I'll turn it over to you for some opening comments. And at the end of the opening comments, Arvind, can you also tell us how many books are behind David?

Well, let me actually start off with that. I think he's got about 80 books back there. And he has read all of them.

Okay. No, he's wearing all of them.

[indiscernible] in terms of the house, Arvind. But that's okay.

Well, listen, Umer, thanks for inviting us to your conference. I think this -- it's just a great custom. We certainly enjoy being at your conference. Before we jump into your questions, I thought I would make some very brief introductory comments so we can highlight our broader investment fundamentals. And I would start off with a message on execution. Even though we are in the midst of a pandemic, we delivered 9% revenue growth, 14% EPS growth through the first 9 months of the year. And of course, the revenue growth has been backed by strong volume growth. If you look at Q3 alone, the volume growth was 18%, and that's being driven, of course, by many of our product growth drivers, including Repatha, Otezla, Prolia. And let's not forget our biosimilars business, which generated nearly $0.5 billion in revenues in the most recent quarter. We have also made very good progress in the Asia Pacific region with our revenues being on track to exceed $1 billion for the first time this year. Good discipline on expense management, and this is at the same time as making commercial and pipeline-related investments. And we have also made some very good progress in advancing some key pipeline molecules, including sotorasib and tezepelumab. And the pivotal data for both of these compounds, as you know, is going to be presented at scientific congresses within the first half of next year. We are going to host an Investor/Analyst event right after ASH, the big hematology meeting, to highlight progress that we are making with our white platform. And let me just conclude my introductory comments by noting that we have a strong balance sheet and will continue to take a disciplined approach to capital allocation. So with that, let me turn it over to you, Umer, and we'll be happy to address any questions that you might have.

Okay. Great. We'll jump right into a lot of topics, but let me just start by saying, Arvind, I remember dating back to my very first days on the job, you're the only survivor who I have continued to work with in the same role for the last 11 years, dating back to my Mark Schoenebaum days. So always a pleasure to work with you, Arvind. But I will say, Arvind, part of the reason I say that is because I remember distinctly actually, having a conversation with you where my mentor, Mark, was very upset at me for asking you this was, there wasn't necessarily a clear track for pipeline at the time at Amgen. And there was a lot more about, "Hey, we have these erosions modeled in. We're going to have slightly less erosion than that, and there's all this capital allocation, which drives the upside " And I recall, always asking myself, well, I don't necessarily find that to be the most sort of exciting next-gen products to be working on, considering the history of the company. And I feel like there's been a big 180 sitting here today with several of the progresses that are made on the pipeline. So my question to you is -- maybe the question is less for David and Arvind, more for Peter. Peter, how do you think about the ongoing pace of investment and expense management? In the backdrop being there's now real R&D productivity kicking in. But on the flip side, some of the key things I've worked historically were margin expansion and capital allocation. So how are you balancing the 2?

That's it. Look, that is a great question, Umer. And it's one we welcome because capital allocation at Amgen continues to be a forethought, not an afterthought. And I think it's important, let's cover 2 parts of your question. First, on operating margin. Certainly, the company back in 2014 when it initiated the Amgen Full Potential transformation started, and I think the operating margin at that point was perhaps in the mid- to high 30s. It worked its way up into the low 50s by 2018. Certainly, the company has very strong transformational muscle, has strong productivity muscle. It's a very efficient enterprise. And we expect to continue that. On the basis of any number of financial metrics over the coming years, we want to be in that top quartile. And we're going to work to continue to exercise the muscle to do that. So that's first and foremost. Secondly, in terms of capital allocation, it's a very important question. We're very clear on this. We want to be predictable, and we want to be consistent. First and foremost, our capital allocation goes to internal innovation. Secondly, our capital allocation goes to our capital expenditures to build our business and enable the manufacturing and the technology around it. We're finishing, for example, a plant in Providence, Rhode Island. We call it our next-gen [ Biogen ] Plant 2.0 manufacturing in the future. And that will be up licensed and running in the first quarter of 2022. We've been able to build that in about half the time and about half the cost of one of our older plants. Secondly, that's going to operate at probably less than half the OpEx that our other older plants operate at. So we allocate, secondly, to CapEx. The other part of CapEx right now, as with all companies, we understand digitization is important. So we're allocating capital to that. We'll continue to do that to digitize the business. Third, we're going to continue to be a company that's balanced between internal innovation and external innovation. And external innovation requires business development. We have an outstanding team led by Dave Piacquad who's been doing that for well over 3 decades. We'll continue to allocate capital to that. We'll do it patiently, prudently, precisely and promptly, as I like to say the 4 Ps, and we'll be thoughtful about it. But then we're going to continue to return capital to shareholders. We're going to grow that dividend. It's $1.60 per share per quarter right now. It was up about 10% from 2019. And then finally, we're going to continue to repurchase shares. We indicated this year, we'll be at the lower end of the $3 billion to $5 billion share repurchase range. And all of that is built on an efficient capital structure that generates an optimized WACC. And that maintains our strong credit ratings because our job is to enable the firepower for Dr. Reese and our team to innovate internally and to allow us to build out our plants and increase our capacity in a very thoughtful way. And third, to get our business development done. So we'll continue to interrogate and prosecute all of that. We think it's really important to be kind of first-in-class in terms of enablement and efficiency and transformation muscle. And so we're going to continue to push forward on that.

So Peter, if I take what you just said, and I have found those -- the word you used predictable and consistent to be very important, because if I take that predictable and consistent track record on capital allocation -- and in the past, I've discussed with Arvind, the pace of repurchases that have happened, et cetera. If I just mathematically plot that out going forward, theoretically, you guys can produce $3 to $4 in earnings power from the 2020 run rate, off of just capital allocation alone in the next 5 years, which effectively gets you to consensus like numbers without even doing very much from the pipeline perspective, and that's where all the programs Dave is working on kick in. So is that a view you disagree with?

Well, as you know, Umer, we don't provide long-term guidance. So in terms of our capital allocation, what I like to say and predictable [Audio Gap] look at Amgen even over 10 years. We started the dividend about 10 years ago. We've been very predictable on that. We started the transformation. We were predictable on that in terms of what we're going to get to in terms of our operating margin and the creation of capital for internal and external opportunities. So I think that's the most important way to think about it. And that the allocations to shareholders after internal innovation, after our capital expenditures, after our external innovation to get down to getting capital back to our shareholders through the dividends and through share repurchases, will continue to be important to us. And so while we don't go forward and project into the future what our repurchases would be, we think history is always a good indicator, and it's important to look at that. And that does contribute to what you've pointed out, which is predictable, consistent, thoughtful about that.

Got it. Okay. So speaking of big drivers of EPS from here, so the capital allocation is one piece. There's one piece that David sits on, which is the KRAS launch, in my opinion, which could possibly be north of 85%, 90% operating margin type of launch. So without getting into the margin profile of the launch, let me ask you this, David, how do you think about the potential -- I'm not talking about the sales number. I'm just talking about the size of prevalence pool for EGFR lung and KRAS lung is probably not that different. So how do you think about median duration of therapy in a first-line setting for the KRAS versus what we know about EGFRs?

Yes. So I think, first, if we step back and look at the current monotherapy data that we've talked about publicly, that we presented, updated Phase I data at ESMO, we've publicly announced the Phase II data looked good. You'll have a look at that at the end of January at the World Congress of Lung Cancer meeting where we'll have an updated data cut and additional biomarker data. We feel very good about that in the second, third, fourth line settings. And then the question is, of course, where do you go from there? As you point out, the size of the market is probably roughly comparable to EGFR, maybe a little smaller, but bigger than the ALK market and certainly bigger than [ RAS1 ] or some of the other specific mutations for which targeted therapies are available. One thing I would point out, based on the biomarker data that we presented at ESMO is that so far, we have not seen really co-mutations travel with one another in terms of other actionable targets. So -- and in fact, you don't really have EGFR mutations. You don't have ALK mutations. They're exceedingly rare in the setting of a KRAS G12C mutation. They're all -- biologically, they're almost mutually exclusive. And that's something that's been predicted for some time for various reasons. So as we move into the first line, the question is, can you do as well as PD-1 checkpoint inhibitors alone or in combination with chemotherapy, and we're enrolling to that combination right now. Over the first half of next year, we'll be talking about those data. I think that's really important as we think about moving into earlier lines of therapy. And then the question is, are there some sets of patients, for example, who simply aren't eligible for checkpoint inhibitors. For example, they have underlying autoimmune diseases or other contraindications. And so is there utility? And then to the final part of your question, what does duration of response, progression-free survival look like in that population. I think these are very important questions for us to address over the course of 2021. I will point out that we are quite pleased with the duration of response data that we presented at ESMO from the Phase I data. And again, we'll have updated data from the Phase II study just next month now, but we feel very good about where we are there.

Bo, do you want to touch up on the point on PD-1 combos we were discussing from other small molecules?

Sure. So we look at the EGFR and ALK inhibitors. And from that, actually, the PD-1, PDL1 just doesn't add much on the durability. In contrast, some of the combos do have some unexpected toxicity. So I just want to have your thoughts on -- we think that -- you mentioned that 30% to 40% of benefit is what you're looking for. Just want to hear your thoughts there.

David, basically, the underlying question is, is there higher risk in PD-1 combination than market is understanding with the small molecule?

Yes. I think that's one of the core questions that we have to answer now. We know from some of our preclinical work that you get upregulation of MH -- probably upregulation of Class I MHC. There appears to be immune infiltration that is precipitated by treatment with a G12C inhibitor. And the question then is, does that give you additional bang for the buck. And then what is the overall safety, as you pointed out, in some of the targeted therapies have not played well with checkpoint inhibitors. I think that's one of the key questions that we have to address going forward and/or that they haven't seen a robust additivity or synergy in terms of efficacy. So all of these questions are to be answered, but I think you're spot on in terms of what the core questions are, right now.

David, outside of TKIs and RCC, has any small molecule combined well with PD-1, I guess [indiscernible] could be one.

That's probably the best example right now, probably the best example right now. But this is also -- it's a different target. It's not a receptor tyrosine kinase. It's a RAS molecule. And there's, I think, a lot of biology to understand here.

David, one of the questions I've had is, I think there's a real likelihood that -- we don't know, it's possible though, that maybe there is not a synergy for PD-1 with KRAS. So that scenario is technically still on the table. If that were to happen, are you thinking about designing your PD-1 trial in a way which answers the clinical question? So what I'm really getting at is, if someone is KRAS positive upon diagnosis, should they be taking KRAS first, maybe plus chemo and then followed by PD-1 rather than vice versa? Because I think that will be highly irrelevant from your perspective.

Yes. No, I think that's a great question, and that's one that we're very interested in exploring, which is, if you don't have direct combinations, is some form of sequential therapy actually the optimal way to treat these patients with drugs that have orthogonal or nonoverlapping mechanisms of action? And then if so, what is, in fact, the ideal combination? Which molecule comes first? Which molecules come subsequently? So those are questions we're really interested in exploring and we'll be exploring in the clinic. But sequential therapy, I think, is also something to keep an eye on.

Got it. I noticed Mirati is doing treating post progression. It's something Arvind and I discussed at length when the triple data came out from Mirati. Were you guys aware of that? And is that something of interest you? Do you think they can put up a better PFS because of that?

Yes. We -- yes, I don't know. It's compared to what?

Sorry, [indiscernible] it will be an impact [indiscernible] .

Yes, that's one of the questions. One of the questions that we're also interested in addressing, for example, is that progression, if you add then another agent, can you recapture response prolonged progression-free survival, prolonged overall survival? And there's some evidence with retreatment with other targeted therapies, for example, after holidays. So I think all of these questions are ones also of interest in the clinic. One thing to think about is we've been in the clinic just a little over 2 years. It took 40 years to get the first inhibitors so there's still a little bit of biology to explore here. We have...

Quality problem.

Yes, these are -- yes, as a drug developer, these are the problems you love to have.

That's great. My last one on KRAS, because there's so much more to talk about. I got a sense listening on the last earnings call that there was a bit of a shift in body language on colorectal. The data we had seen previously would not suggest to me that Amgen would come out on an earnings call and point people to a colorectal cohort coming up in first half. So I was curious, do you -- are you familiar with more data on colorectal than what's been disclosed? And was that the intent?

No. I wouldn't overread that. I wouldn't overread that at all. We've fully enrolled Phase II trial in colorectal cancer in monotherapy. Again, these are advanced patients with multiple prior lines of therapy. It's important to keep in mind the sort of baseline here. Currently available drugs have response rates on the order of 1% or 2%, with progression-free survivals reported on the order of a couple of months in a number of trials. So it's a relatively low bar. I think when we see the Phase II data, that really tells us since there a monotherapy pathway forward or is combination therapy in colorectal cancer likely to be the answer. And of course, we're aggressively prosecuting the relevant combinations in that indication in parallel.

Got it. Okay. So maybe moving past KRAS then. I was hosting a European pharma company, no name, you can probably figure it out, this morning. And they've mentioned a very interesting point I had not thought about, which is from your asthma partners' perspective, when we think about high eosinophilic and low eosinophilic market in asthma, on the high eosinophilic side, they already have a fully owned asset Astra does. On the low eosinophilic side, obviously, it will be TSLP. Is there an incentive in place which incentivizes them to maybe not focus so much on TSLP on the high EO side? And how does Amgen's commercial effort tie into this broader TSLP launch?

Yes. Let me -- I can start there, Peter, you may want to jump in or, Arvind, to add a few words. First of all, I'd say we're very, very happy with the partnership. It's worked extremely well. One way to put this into perspective is that there are roughly 1 million patients in the United States alone with severe asthma who are candidates for biologic therapy. They split between the high eosinophil and low eosinophil categories. Roughly 60% or so have eosinophil counts that are 300 or less. There's no formal scientific definition of low eosinophil versus high eosinophil, but 300 is often used as a cut point. And as we indicated in the press release for the NAVIGATOR study with which we're thrilled, we showed efficacy across that range of patients with high eosinophil or low eosinophil. And in particular, I'd call out the group of patients with eosinophil counts less than 150, which accounts for probably 30% or so of patients with severe asthma, where no drug has shown these kinds of results. And in fact, if you look at the under 300 category, it's the first drug that has shown statistically significant, clinically meaningful improvements in outcomes such as exacerbation rates and some of the other biologic parameters. So tezepelumab is a drug we're very, very keen on. And I think my view is that this should be an important drug for patients with asthma.

Got it. Got it.

I have nothing more to add there, Dave. I think you've covered that well. I think just the fact that tezepelumab has a potential in this broad population of severe asthmatics, regardless or irrespective of their eosinophilia type, I think that's really one of the key advantages that we can potentially bring to the market with this product.

Got it. Arvind, are you guys co-commercializing with Astra?

Maybe I'll jump in here. We're going to commercialize in the United States and Canada. They'll do it solely in the rest of the world. We'll record sales in the U.S. that recorded elsewhere. And we'll take our profit share in through other revenue, and we'd get a single-digit inventory shift royalty on this. So this is a great relationship. We're -- as Dave and Arvind have both said, we're very pleased with it. We're looking for...

They're not even commercializing in the U.S.

It's joint.

Oh, it's joint in U.S.

Yes, joint U.S. and Canada.

And they will do ex U.S. by themselves?

Correct. Ex U.S. and Canada by themselves.

Cool. Okay. Got it. Okay. Got it. Makes sense. Okay. I don't know if that's -- it's a fair question to ask right now. This might be something you guys might address at some point. But would you think of the KRAS opportunity as being equivalent to or smaller or bigger than the TSLP opportunity? Or let me add a third one in there, omecamtiv. No? It's a bad one.

No. No, we don't go long-term guidance. Let's just say, as Dr. Reese said, we're very excited about teze and KRAS and 12 -- G12C. So we're very confident. And we've allocated plenty of capital to [ Murdo ] to make sure we get those launches off successfully, and we're looking forward to working through all that.

Yes. And I mean I'm incredibly proud of the R&D organization here. These are 2 first-in-class novel mechanism of action drugs that we've brought along. And so it was a big year for us to deliver those data.

Okay. Excellent. I want to go to oncology in just a minute, but I feel like there's an asset I've been personally sleeping on despite a lot of the work our team and I did on the RNAi space, the Lp(a). And John Shutter from Arvind's team, he sent me the presentation recently. And it looks like -- I mean, look, there's -- it looks like not only is it pretty active. There's also a very long and durable sort of response to that. But I guess my question is, how should we think about it relative to some of the other Lp(a)s in development? And where would it tie in versus a long-acting PCSK9? And also knowing that PCSK9s have a meaningful Lp(a) reduction as well.

Yes. No, it's a great question. This is a program I'm very interested in, AMG 890, which is a small, as you mentioned, a small interfering RNA that targets Lp(a). Maybe just for the benefit of everyone who's listening who may not be familiar with Lp(a). A few words on that as a target. There's a wealth of genetic and epidemiologic evidence that ties elevated levels of Lp(a) to atherosclerotic cardiovascular disease. And there are a couple of important features of that risk. One, it's independent of LDL cholesterol and purely independent as a variable, actually. Number two, it doesn't appear to be modifiable. So diet, exercise, all the usual things that we would do, for example, to control cholesterol levels outside of pharmacology have no effect on LP(a) levels, which appear to be basically genetically fixed. And so right now, the only way we know to test the hypothesis of whether lowering Lp(a) improves cardiovascular outcomes is through pharmacologic manipulation. And as you mentioned, a few weeks ago, we presented the Phase I data from AMG 890 at the American Heart Association meeting showing that doses of 9 milligrams or above, you got 90% or greater reductions in Lp(a) levels that lasted for up to 3 to 6 months. So we've now moved on and we're actually briskly enrolling in a Phase II trial and are interested in carrying that drug forward. It could be a great complement to Repatha in the cardiovascular franchise. And Lp(a), again, as a driver independent of LDL is often individuals show up in their mid-40s, early 50s with significant heart disease, but don't have an elevated cholesterol. And these are the patients we think could really potentially benefit from this sort of drug. I'll add in closing that we're also using extensively decode genetics and some of our other omec approaches here to inform the development program. And we really feel that, that's an important part of the future broadly for drug development, but this will be a flagship program to really start to, early in development, tease out subsets of patients who really will derive the greatest benefit.

Makes sense. Makes sense. Makes sense. Excellent. Bo, shall we turn to BiTEs, if that's okay?

Sure. Sure.

Okay. Go ahead, Bob.

Yes. So maybe could you first give us a quick thoughts on among the BiTE programs, which ones are you the most excited about?

Yes. I mean I think there are -- the ones we're presenting data on this year are really in the half-life extended program, the ones that we have great interest in. At ESMO, we presented, as you're aware, data on AMG 160, which targets PSMA in advanced prostate cancer. We're continuing to dose optimize there. I'm quite pleased with what we're seeing in terms of efficacy and our ability to manage adverse events such as cytokine release syndrome. So I think that's a program really to keep an eye on. And then the second one in solid tumors, AMG 757, which was just presented a couple of weeks ago at the SITC meeting, the Society for Immunotherapy of Cancer meeting, that targets DLL3 in small cell lung cancer. We showed responses in patients with heavily pretreated disease. This is a disease where the treatment paradigms really haven't changed much in decades. When I was a fellow on the wards decades ago, the -- more or less the same drugs are being used. Now immunotherapy has not had much of an impact to date in that disease. And so we think there's a real opportunity there as well. And so those are 2 programs I'd really keep an eye on over the course of 2021, and we'll be happy to share more data as they emerge. And then in a matter of days, we're going to be sharing some of the data from AMG 701, which is the half-life extended BiTE targeting BCMA in myeloma as you may have seen in the abstracts in the highest dose cohort. Included in the abstract, 5 of 6 patients with responses, and we're presenting more fulsome data, again, just in a matter of days at ASH. So those are, I think, the programs to keep an eye on right now. And while they all remain in dose escalation, I'm optimistic about what we're seeing in getting one or more of those to really go and go aggressively.

Right. David, I have to admit, as I've looked at the data on the BiTEs broadly, I find that there was a lot of chatter on BCMA, and there's a fair amount of chatter on PSMA, but very low on DLL3, even though the way I see the data is on PSMA, the CRS was as much as 90% and a lot of that was Grade 3, 4, whereas the CRS was much, much, much lower on DLL3. Also the responses on DLL3 and small cell look much lower. So to me, it looked like DLL3 was probably the best one of all of these from a clinical profile perspective. Would you agree with that?

I would say not necessarily. I would say I love both these children. And I would say with further dose optimization in the PSMA program, I think we're -- we have line of sight to making that a very manageable toxicity profile. And I think there's a real opportunity to also enhance efficacy there. And with DLL3, as you mentioned, CRS was less of an issue. I think that illustrates a broader point that I also like to emphasize, which is, the toxicity that you see with bispecifics is very target dependent. Now CRS is, of course, a general phenomenon, but how it manifests itself how severe it maybe is very, very target dependent. And that's one thing that I think just our extensive experience and having treated over 3,000 patients now with bispecifics and the extensive preclinical programs that we have in place to guide us in the clinic can really help.

Got it. David, I guess said another way, do you envision a pace and rapidity and a path to pivotal trial for DLL3, which is not too different than KRAS, given the signal we kind of already have at this point?

Yes. I think once we settle on a dose that we feel is the take forward dose, then I think in both these programs, the question we'd be asking is, okay, what is the registration path right now? That's absolutely the kind of thinking that we would put into place. Bring into place.

Got it. How many patients can you dose per month? Because I find that with bispecific things always go much lower than you think. I mean, I know at the 3 mg dose, which is -- was of high interest from the early data, it was 3 out of 7 patients were responders. How many patients of data do you need to see to be in a potential pivotal cohort? And when would -- when could that be?

I think certainly over the first half of next year, potentially earlier, depending on what breaks we could fix on what the go-forward doses may be for these molecules. And then I think if you're seeing your response rates on the order of what you've described and you can demonstrate durability, that's probably a viable clinical profile.

So if, let's say, come ASCO next year, we're seeing response rates hold up the way they are -- or even in the same ballpark broadly, you're suggesting we could be entering a pivotal cohort, which could be a potentially fileable study given what we saw in lurbinectedin, for example.

Yes. I mean I don't want to get ahead of myself and speak for the regulators, of course, but that's exactly the sort of thinking that we would be taking forward in the development program, which is what is the pivotal trial.

And there's no gating factor from a recruitment or manufacturing perspective to, let's say, dose 100 patients in a pivotal trial for second half?

No. I remember the BiTE, even the half-life extended BiTEs are given in still small doses. And so we can do manufacturing runs and supply a large, large number of patients. So drug supply is not an issue at all.

Got it. So theoretically, then, David, it sounds like without having any guidance per se on it, but it sounds like it may not be pivotal, but you could have a Phase II monotherapy refractory small cell cohort wrapped up in 2022.

Yes, that would be a reasonable aspiration. Of course, it all depends on the data and how dose escalation plays out, but it's certainly moving along with that kind of tempo. I wouldn't want you to put words in my mouth guaranteeing that. But the program is moving quite quickly. We're not having trouble recruiting patients because, again, there are such so few treatment options for these patients.

Got it. And remind me that is the DLL3 only recruiting DLL3? Is there a biomarker and what percentage of small cell is that?

We're not screening for DLL3. We're looking at that, but it's nearly ubiquitously expressed in these tumors. And we believe expressed at levels that are more than adequate for bite engagement. Our best guess is that you need a relatively small number of targets on a tumor cell for effective BiTE engagement and T cell activation.

Okay. Bo, anything we missed?

On the PSMA program, I'm curious what are your thoughts on the heart toxicity like afib and the tachycardia and [indiscernible]

Yes. I think those were all -- those aren't direct cardiac toxicity. Those are manifestations of cytokine release syndrome. And as we're going forward, I think as we manage CRS, you should see those sorts of side effects fall away. That would be my guess. That's -- we have to demonstrate that, but I'm actually pretty comfortable that we'll be able to manage that. And again, I think a lot of these things are manifestations of underlying cytokine release.

And we have a question from investors. And could you provide an update on the gluten [indiscernible] study, please?

Yes. So this is -- for those who don't know, AMG 714, that's a monoclonal targeting interleukin 15, which was believed to be one of the drivers of celiac disease. So with our partners that has entered Phase II testing, potentially pivotal Phase II testing. And so we'll provide guidance as that trial moves along, but a lot of interest there because there's quite a bit of need in those patients who've got refractory celiac disease, but actively enrolling now.

Got it. Excellent. Anything, Arvind, that you often get asked or something that you're getting asked, which is important, but didn't come up in our conversation today?

Sorry, Umer, I've got the mute on. No, I think you've covered a lot of ground on the pipeline front. I mean in many ways, 2020 was viewed to be a year of pipeline readouts for us. So looking forward to, obviously, this data now being presented in a scientific form next year. I'm glad we had a chance to cover the BiTE platform. I think there's a lot of good work that's being done there. So no, I think you've covered it very well.

Got it. So Peter, can I just sort of conclude this whole session with where we kind of started? It feels to me, as I think about the profile going forward, that there will continue to be discipline on margin expansion, et cetera. We should probably expect additional spend on R&D side, but perhaps not a massive, massive growth on overall OpEx, meaning the expenses will continue to be managed while continuing to invest in pipeline. I just want to make sure I balance that carefully with the investment needed for these launches as well as pipeline.

That's a great balance, Umer. And that's why we were so specific about our OpEx in the fourth quarter. And it was a manifestation of deferrals of expenses in the second and third quarter, in the fourth quarter, and we need to make these investments in our launches, and that's what -- exactly what we're doing. So we'll continue to make those investments. I think you said it really, really well. Disciplined capital allocation. We'll keep going back to that and keep doing that, making sure we're getting that accomplished first and foremost with our internal innovation. We talked about BiTEs. We talked about teze, we talked about soto, we talked about Lp(a). So right around our franchise, and we're going to continue to enable the funding of that, make sure that we get that done really, really well through very thoughtful and strong transformational and efficiency muscle. We will continue that. I call it, continuous innovation. And half of the portfolio over any number of years will come internally and roughly half externally as has been historically. We're going to continue to be reliable executors in terms of getting to that kind of the top quartile of performance in terms of our financial metrics at which operating margin is arguably one of the most critical, if not the most critical. And then we didn't talk much today about our integration capabilities and how we used $13.4 billion of our shareholders' money just slightly over a year ago to acquire Otezla. And that integration was on time, on track. And we never take that for granted. That was a big allocation of capital. The collaboration with BeiGene is working really well. I think Arvind alluded to earlier that we expect to go through $1 billion of revenue in our Japan and China this year. And we're going to continue to work hard in the second and third largest pharmaceutical markets in the world. So it's continuous innovation, it's reliable execution, it's timely integration and disciplined capital allocation. And we're going to continue that going forward and what we've been doing historically. And we're really delighted to be able to have an opportunity to share that with you and Bo and those who are interested in the company today.

Fantastic. Thank you guys so much for joining us. David, thank you so much for bearing with our random questions. Arvind, thank you for your time as always.

It's been a pleasure.

Thank you.

David, I will ask you how many books there are behind you next time.

Okay. I'll give you a rough tally.

Sounds great. Thank you again.

More than 80 but less than 1,000.

I'm going to go over on the 1,000, Arvind. Thank you, gentlemen. Thank you again.

All right. Take care.

Thanks a lot, Umer.