Eli Lilly and Company (LLY) Earnings Call Transcript & Summary

Good morning, everybody. I'm Chris Schott at JPMorgan. And it's my pleasure to be hosting a fireside chat with Eli Lilly's Chairman and CEO, Dave Ricks, today. Dave, thanks for joining us. I know you always bring some great news to the conference, and certainly, the Alzheimer's update yesterday really kicked off the conference on a positive foot. I thought maybe just to open up, if you want to just make some general comments around your priorities for Lilly in 2021. Obviously, a lot of topics we can dig into from there but make some opening comments, and we'll go from there.

Absolutely. Well, thanks for having us, Chris, and look forward to talking about Alzheimer's later maybe. But 2020 was a kind of remarkable year in a lot of ways, and it sets up what we're focused on in '21, which is -- the first thing is we're in the middle of a strong growth period for the company, double-digit volume growth and high single-digit revenue growth throughout last -- first 3 quarters of last year. And we just need to continue to execute against that with the opportunities we have. We have a broad and diverse portfolio of launched medicines that are driving that growth. We expect this year to get some -- to something like 60% of our revenue in that newest slate of products, which is a fresh lineup with their own competitive classes, and we need to execute well. So that's the first priority. And that's against the headwind of the pandemic still going on. We still see suppressed visitation in doctors' offices and volumes are down a bit from where IMS and those market observers had guessed they would be. So we need to fight through that with share and class growth. That's top of mind. We are expecting the second half for that to clear up quite a bit. So in particular, the first half, we got to execute through the choppiness. The second thing is really making sure we stay focused. We have a great productivity and margin expansion story. How do we get -- because we're operating in a focused set of therapeutic areas and we're adding products on top of infrastructure, most of that margin expansion we've been talking about is coming through the SG&A line by just putting more products in the bag of people selling to the same doctors. I think if we're -- as the company grows and we do well, the appetite to expand is always there. And I think that there's a fine balance between taking advantage of opportunities and being focused on what you need to deliver. So that's -- kind of drives the productivity story for us. But I think, actually, the business decision is to stay focused. And then finally, I think that last year, there was a lot of discussion with investors on -- we feel good about everything until perhaps Trulicity's patent expiry in -- later 3 or 4 years of this decade. I'll take that. Whenever you're talking to investors about something 6 or 7 years from now, it probably means you've got a good future. And I think just in the last 60 days, and we can talk about this today, I'm sure we will, we've seen some of the seeds of growth that will drive performance beyond that important date for Trulicity, like the tirzepatide results, like the BTK inhibitor from LOXO and like donanemab, which we announced just yesterday. So sustainability for the long term, the R&D engine is a priority. And we're very interested in new modalities and many other projects beyond those 3 I mentioned that could really contribute beyond what's right in front of us today. So those are the top things. Maybe I'll just add one more, which is I think all companies are thinking about what do we just learn by operating the way we did during this pandemic that could be useful going forward, things like how we run clinical trials with much more remote monitoring and digital interactions with sites, how we sell, how we run our company. So that's also a priority this year.

Yes. On that last topic, do you think there will be ultimate -- whether it's cost savings or just accelerated R&D time lines, I mean do you think that there are going to be some tangible benefits when all is said and done? Or do you think we kind of drift back to the way business was done in the past as we get some time between us and the pandemic over the next few years?

I think there will be permanent changes in a few ways. Probably the 2 most important for the industry are how we run big, primary care, late-stage studies where you have hundreds of sites. I think we learned to do that and surprised ourselves at how effective it was to do virtual support of those sites and quality monitoring and data gathering and so forth. And even with sites to patients, the same. Really, the dropout rates during the pandemic surprised us, how strong they were. I mean we announced the tirzepatide results. That was a question we had on the call. Really, I mean there's a small bump you can see early April last year. But really, we sustained enrollment in all our programs incredibly well despite huge problems, transportation and the health care system. And so that teaches us that we can do it with a lighter touch. We can do it. Maybe reach new populations that were harder to reach and maybe even enroll studies a little faster if we do digital support. The other big one is how we go to market. And the chronic problem we face as an industry is, in primary care especially, every year, 2% or 3% of the physician groups get bought by a hospital system that bans pharmaceutical interactions or just decide we're not going to see reps anymore. But they need information still. They need support. And that field is growing rapidly. It grew a lot last year. I think our virtual contacts were up like 1,000% over the prior year in the U.S. internationally, a little bit less impressive growth, but they already had quite a bit, particularly in Asian markets. So that will become a new feature of all our go-to market. And maybe we can reach those physicians who wouldn't see our sales representatives much better if we use the technology.

Yes. Maybe just last big picture one for me is on pricing. It seems like we've seen modest price erosion across Lilly's portfolio the last couple of years. Do you see anything on the horizon that could significantly shift that price dynamic either positively? I think people are probably more concerned about negatively. But just -- do you think we're having this trend where a couple of percent erosion a year is a pretty good proxy or are there other dynamics we need to be watching on that front?

Yes. And I think in our guidance, we guided toward that. We expect more of the same. The challenge, I think, for investors is it doesn't always come smoothly across quarters, right, that there's timing in the benefit design of Part D that causes uncertainty in our projections, that there's various pushes and pulls in the commercial market on rates and that product rollouts have different adoption across segments, Medicaid being a notable one, where when you start to get a lot of Medicaid coverage, you're growing profitable volume, but it's at a lower rate. Those things don't happen evenly. And so that, I think, gives us a lot to talk about on quarterly earnings calls. But the macro trend over the last 3 years where we've had really markedly reduced list price changes across our portfolio, particularly in diabetes, is 0 on insulin since 2017 and mid-single-digit across most other products, where before it might have been high single digit or even double digit. But that's -- we're in that mode, I think, as far as the eye can see. There's always the policy discussion about is there some threshold event policy-wise. We talked about that on Friday at your conference. I don't see a big risk of that in the short term. And so for the next several years, more of the same for us. Those pushes and pulls, quarters may look better. I'm surprised people are worse, but the long-term trend is that single-digit price headwind. Fortunately, we've got a huge volume opportunity, and we're growing volume in major markets in the mid-teens. So that's overcoming, and then some, that price headwind.

Absolutely. Maybe just digging into the portfolio. Obviously, big news in the week on the Alzheimer's update yesterday. Just to kick off, maybe just help us put that data into context and talk a little bit about your overall excitement with donanemab and what you've seen so far with the data.

Yes. Well, maybe just to step back on how we got to running that study to begin with. And around here, we've been involved in running Alzheimer's studies for some time, as you know, and kind of rather dissatisfying exercise to this point. I think probably no entity has more experience in looking at failed Alzheimer's trials and trying to improve the framework by which we try to ascertain a benefit for these patients. Then on the other side, you also have the drug itself. So let me talk about those 2 things. The first is in terms of how to conduct studies. The instruments we use to measure this disease are noisy. They have a lot of radar variation. Some of them are long, and so there's even [ ray T ] fatigue in doing these studies. And I think we used to believe that the way to run studies was to overcome those things by scale. So one of the criticisms, I think, of the study is it's small. Actually, I think that's a strength. To show a benefit in a small study is remarkable because the path in the past was to power up based on getting 800 patients in an arm or whatever to overcome that variability. You can do that, and we got close even on EXPEDITION3 by having almost 1,000 people in each arm. But that doesn't mean you make a clinically important drug, right? You can detect the difference, but the difference could be quite small. And even by powering up, you create new variability because you have more and more sites that create points of difference across many countries. These instruments aren't all translated into local languages. There's lots of challenges with that powering up strategy. So we moved away from that. And we said we need to do smaller studies that have a higher threshold of effect. The second thing is using clinical ratings of disease performance is also hugely variable. So when you have an Alzheimer's diagnosis from primary care and send it to a specialist, something like 1/3 of the time, the specialist will change that diagnosis to some other dementia. So -- and even between special centers, you'll get changes. So clinically, it's difficult to ascertain pathology. You need to use pathologic tools like PET scanning. So years ago, we started amyloid scanning everyone. That's now standard. In this study, we added tau PET because tau is a key marker of symptom progression. And so if we wanted early people with disease, we needed kind of that Goldilocks tau scenario of, not 0, because they might have -- there are people with no -- with a lot of amyloid and no tau who never get symptoms and also not too much tau because there are people with a lot of cognitive reserve who have a lot of tau and a lot of amyloid, and they tend to fall off a cliff symptomatically at some point. And those are those rapid progressors a lot of other companies have talked about. The way to eliminate that, in our mind, was to bracket this tau group, and we did that in this study. So with the right population, then the question is, what's the right drug? And in our mind, we thought the amyloid hypothesis needed to be really tested. We thought this was going to be the study to test it because if plaques matter, then you need to get rid of them. And you need -- the sooner you do that, the better. And the more robustly you do that, the better. And that's what donanemab does. So we published our Phase I data. But in this study, I think we also spoke about the amount of clearance is more than half the people at the end of the study had no measurable plaque. So they're below normal or at the normal 24 centiloids or below. And I'll share with you here qualitatively, we're going to publish this data soon, that actually at the first measurable time point, a significant portion of patients had no measurable plaque. So this drug works not just -- a lot of people publish those endpoint numbers, and we're getting to 80-plus-percent reduction on average, and that's good, better than any other drug we've seen. But the story is even better than that for donanemab because at the first measurable time point, in some cases, you're already below the plaque at the last measurable time point for any of these other plaque-clearing antibodies. So deep and rapid clearance matters. And I think that showed up in a relatively small study. We overcame the scale of study with precision on who got in and a very potent and effective plaque-clearing drug and hits that [ sig ] on the primary endpoint, which is, I think, the first study ever to do that.

Yes, yes. What do we think about in terms of next steps for a program like this? So I guess the -- one of the big questions is, what do you do with this data? Do we need another study? What's the FDA thinking? How are you envisioning kind of where we go from here?

Well, of course, the FDA wants to see this data, and we want to show it to them. They have policy levers to rapidly approve drugs. I hope they would look at that here. Of course, that's mostly not been applied in neuroscience. That's been applied in other disease settings. But that's possible. That's their judgment to make, not ours. That's why, yesterday, we said we're committed to do the second study anyway. We think, first of all, these results are strong and reproducible, and we plan to reproduce them. In fact, we made a bet I'm glad we made last summer by launching a replication study. We'll probably make a few tweaks to it based on what we saw here and call that a Phase III or a pivotal or whatever, a second pivotal. But the standard of the FDA is typically -- is 2 well-controlled studies with 0.05 or less on the primary, and we have 1. So that's -- the second one could serve us that purpose. Of course, we'll advocate to make access available earlier, but that's the FDA's call.

And when can we think about seeing more of this data published to get a better sense of not just the primary but some of the secondary trends that you were referring to in the press release?

Yes. Quite soon, actually. So we're excited to say today that we'll be presenting at AD/PD, which I believe is the week of March 9, and so that's relatively soon. We have planned a simultaneous scientific publication in a major journal at that time, so pretty soon. And I think really investors need to see that we had a limited press release in terms of what we disclosed because we expected a rapid disclosure in a scientific forum, which is the right place to get into the detail. But I think investors will continue to be impressed with the data, not just hitting the primary, which is this ADAS metric. Just to clear that up, that's just the numerical addition of 2 well-established metrics for ADAS-Cog13 as well as the iADL. So it's not anything too esoteric. We'll show those subcomponents, which are good, CDR Sum of Boxes, MMSE. So those 4 pieces, and then 2 of them add to make the primary, are the key secondaries. But in addition to that, just to point out, Chris, we also are using, as a key secondary, this disease progression model. So we look at all those metrics not just at the final point but over the course of the study, kind of an area under the curve analysis, which was a key secondary as well. And all this is showing to us very consistent kind of every time point every measure trending to a drug, and in the case of disease progression and in many of those metrics, hitting more than just the end time point statistical significance So it's a strong set of data. We're excited to disclose that.

Yes. A couple of quick ones. Just on the timing, if you do need a second study, when should we think about a time line around a second study if that's ultimately required?

Well, it's hard to nail that down exactly. I think we've got data out there in the clinicaltrials.gov world, but that will be adapted. We'll probably make a few tweaks to the study. We may add a few more patients for safety purposes. There's this high tau PET arm that's included in the current design, not knowing whether that mattered when we stratify the tau entry in this current study. So we'll look at all those parameters, probably not change it too much, but may adapt it a little bit. But really, the driver will be recruitment, which we hope will pick up rapidly right now. There's a lot of interest in the study. So hopefully, we can recruit that in the coming months. I think right now, it has it recruiting through middle of the -- late part of this year. Hopefully, we can improve upon that. And then the duration of the study, we should probably talk about with the regulators, whether 18 months is appropriate or we get away with less or more, what's the right confirmation signal to really have approvability or as -- if there's a conditional approval as confirmation of that conditional approval. So there's a few things to be determined. But I think the clintrials.gov says '23, early '24, something like that. It could come in a bit, but that's not going to change by years.

Sure. And then just last question is, you mentioned this importance of patient selection and kind of getting those patients at the right time frame, how do you see that playing out commercially? Do you think that this is more about proving the hypothesis, and then in a real-world setting, you're using this in a broader population? Or do you see this kind of restricted to those patients who are -- made at the right time in their disease progression with -- I'm trying to get a sense of like how broadly a drug like this could be utilized.

Well, when we looked at screening for these patients, maybe a way to answer the question is, well, how -- are you really looking for a needle in a haystack here with special patients? Or is it a pretty big group? The answer is it's a pretty big group. So if you screen for people with dementia of an Alzheimer's diagnosis, about 30% don't have amyloid. There's some other -- so that needs to get screened out. And that's a onetime amyloid PET test, which is available now. So that's not so difficult, but that would need to happen for our drug or any other amyloid drug. There's no sense in treating someone with anti-amyloid plaque drug with -- who doesn't have amyloid. Then you can look at tau. And as you know, we're working on a blood test that detects tau that we think is pretty accurate. We also have a PET under review, tau PET, that could be available as well. And that could also be a onetime, is what's your entry point tau level? Do you have some is the question. We have this high tau arm in the second study because we don't know the answer to whether this -- the drug would also work in that setting. That would be useful to know, so that maybe we're just worried about people with no tau versus people who have some and cut points and so forth. All that probably leads to something around half of all people with an Alzheimer's diagnosis would fit into those categories. It's pretty good [ sector ].

Yes, absolutely. Well, very exciting data. So I look forward to the March 9 disclosure on that. Lots of other things in the pipeline, too. The other one I just wanted to dig into maybe was on tirzepatide. And maybe before you go to tirzepatide, just talk about the state of the GLP-1 market right now. I know that's been -- it seems like year in, year out, it's been one of these markets that just continues to grow a bit faster than any of us anticipated at least on the investment community. Just your sense of where we are in terms of the -- so that's -- I think the questions I get on Lilly a lot of times are based on Novo versus Lilly and this kind of back and forth. But ultimately, the category, I think, is really the bet we're making. And how much runway do you see for the category as we think about the next 3, 5, longer-term kind of years on -- where can we go here with volumes?

Yes. No. It's the right question. Before we launched Trulicity, I think no one really predicted the market would get to where it is today, which is today about a little less than 40% of patients who failed all-oral options go on to a GLP. And before we launched Trulicity 6 years ago, that was half of that. So that's really been a key part of the growth story for us with Trulicity. GLP market continues to grow. And last year was a tough year. If we look at diabetes treatment in the U.S. last year, HbA1c test as a marker of how people are getting cared for dropped dramatically and particularly in the spring and fall. So we know without that data, doctors can't make therapeutic change decisions. But still, throughout that, GLP is growing over 20% last year. And that number of a little less than 40% of patients who failed oral options is building. In fact, if we look at the new therapy start data, so what's happened at the margin, 1 and 2 are getting GLPs. And I think if you think about the therapeutic choices of a weekly versus a daily insulin -- a weekly GLP or daily insulin, weight loss versus weight gain, hypoglycemic risk versus not, it's really not a close call medically. It's really about habit and cost. And probably 20%, 25% of patients might be lean type 2s or older and insulin is just fine. But most everyone else should be getting a GLP likely. So that would imply, I think, quite a bit of growth left just in that positioning we're in today. Now then we can look at newer technologies. Novo's got the oral treatment. Tirzepatide's coming. I expect that GLPs will begin to grow out of that positioning over time, particularly earlier in the disease course where we have the potential to disrupt the progression of failure that is -- that defines diabetes treatment. Go on to one oral, it wears off after 2 years; go on to a combination of 2; go on to combination of 3. That's very frustrating. It doesn't allow doctors to ever retreat with those therapies again. And patients' pancreases and weight disposition, et cetera, their insulin levels are going in the wrong direction the entire time. So GLPs have the potential to change that. We think tirzepatide, GIP/GLP, could dramatically change that. We need to test that. But that's to me the long term, 3, 4 years out, how could this class double again, triple again? That's how it's -- that can happen.

Great. And then on SURPASS-1, obviously, very positive data we saw from that program. But just your overall take in terms of what you're hoping to see with that, what that means for tirzepatide's role in the market, and just help us put a little bit of context as we think about the remaining SURPASS studies reading out this year, what you're really watching as you think about kind of, I guess, moving this into maybe some more advanced patients over time as well.

Yes. Yes. No, thanks. We're very happy with the SURPASS-1 data. I think it answered 2 key questions on everyone's mind, which was in relatively smaller Phase II studies, we had 2 of them, could we replicate the profound reduction in A1c? And here at the highest dose in SURPASS-1 in relatively new diabetics, half got below 5.7 on A1c. And this is the first time where any study is even measuring that level of reduction. And 5.7 is important because that defines pre-diabetes. So in essence, you become euglycemic. Half the patients become euglycemic on a high dose of tirzepatide. That's a profound switch in how we think about effectiveness, raising the bar dramatically. And of course, the weight effects were good, almost 10 kilos of weight loss at the endpoint on the high dose, which is also -- those are probably interrelated concepts, if you can control weight, you can control early diabetes. And then the second big question is with tolerability, right? So we were in the high to mid-20s on dropout at the high dose in that very first rapid titration study we published back in '18. And through a more staggered titration could we control that as have -- we've seen with GLPs, higher dose of GLPs being managed with titration. And the answer is yes, with 6% dropout at the highest, so more than threefold reduction and dropout. Tolerability looked good in GLP like. So here with the dual-acting, we can get in the same range even at 15 milligrams. So -- and 5 and 10 look even better, maybe you can argue they're even more tolerable than Trulicity. So that's -- those questions, check, check. Remember, this study was in relatively early diabetes. So the numbers themselves in terms of how much placebo-adjusted A1c reduction or weight reduction, this would be probably the least remarkable cohort. The more people progress, theoretically, the bigger the differences could be with tirzepatide. We've got those studies coming against insulin, against the SURPASS-2, which is -- I believe, against Ozempic. And these will be coming in the first half of this year and all key readouts for the program. We'll get that data together and hopefully get it to the FDA here soon.

Yes. And is your sense, I guess, when -- you got this debate around kind of what's the dose we should really be thinking about here. When we think about this drug launching, are you expecting over time that we see most patients at the 10 or the 15? Or do you see -- given how strong data was in 5, that ends up being an important dose? I guess just trying to figure out what -- where you see the most utility, I guess, of the 3 options you're going to have here.

I think we really need to see all the data to define that. And it might be different, Chris, by the population. So here in early diabetes, you could argue that 5 is a pretty good workhorse on A1c. I mean it certainly is going to outperform our GLP and likely others with the dual-acting mechanism if your target is that. If your target is weight reduction and A1c, likely you'll need higher doses. But maybe in early diabetes, we want to be more aggressive with that to reverse the course of the disease or perhaps flatten the curve so that a 45-year-old male who has high BMI and early diabetes, we can keep them healthy for much, much longer. In later populations, we'll have to see what that right mix of dosing. But the main point is we have choices for doctors. And I don't think we'll be likely landing on a spot where we'll say, "Oh, this is the dose everyone should get to." If nothing else, because not everyone requires 20 pounds of weight loss a year, and that's certainly true in elderly patients in more progressed disease, where you do have this leaner type 2, but they can benefit from a -- from tirzepatide, but maybe it's more of an A1c side, so a lower dose might be appropriate there. So anyway, we'll look at the totality of data, make a recommendation to doctors. The good news is we have the range to work with.

Sure. And maybe the last question on tirzepatide. You're running a number of nondiabetes indications, so it's obesity, NASH, et cetera. How large should we think about the market potential for those indications relative to the core diabetes indication?

Yes. I mean it's a large opportunity. Of course, if we drew the Venn diagram, there'll be a lot of overlap between these conditions. Obesity is a big bubble. And we know that chronic obesity has many untoward downstream health effects. Maybe the equivalent of smoking in the 60s, right? I mean you have all this downstream health costs and lifestyle impairment and disability that's caused by it. So reversing obesity has to be a key public health goal. Within that though, it's not always medicalized or understood what those relationships are. We're -- in the last 5 years, I think our understanding of NASH and liver disease, fatty liver disease in obese subjects has grown dramatically. We've known about diabetes for a long time. Obesity is a precursor to type 2 diabetes as we age. And heart failure is another area of emerging interest for us with this mechanism also highly correlated with obesity. So I think what you'll see is -- the way to think about it might be, okay, there's -- almost 100 million Americans are obese. There's a big obese populations in Europe as well. How could we affect the medical consequences of that obesity with tirzepatide? Of course, we'll have the obesity indication. Likely payers will restrict that to places where the health care system has consequence. So what are those bubbles? And can we add to the scientific arguments that treating obesity, because it affects these downstream consequences with something like tirzepatide, which is potent enough and powerful enough to really demonstrate medical outcomes, is worth it. And that's -- we're laying that out now and looking at beyond the normal indications one would think about here because obesity has such a strong impact on health. In the big obesity program, we're looking at a number of secondaries. And within that, that will then define more indications we can go after because we'll be able to observe patients who did well by losing weight in everything from heart conditions to other organ systems.

Yes. Makes sense. Maybe another kind of R&D update. We had a lot of data, you guys, for the last few months here. But pivoting to LOXO-305, where we saw updated data at ASH, I guess just talk about how you see 305 positioned both in the, I guess, BTK-naive market as well as the treatment-experienced patients. I'm turning my hands around kind of with IMBRUVICA pretty well entrenched here, how does the drug data -- the 305 data is great so far. But just what role does this ultimately play in the market? And how do you kind of move, I guess, upstream over time with the asset?

Yes. Good question. I mean -- and to be determined as the data continues to mature. But so far, we're impressed. This started when we bought LOXO as LOXO like targeted therapy play, where we know that some patients who are treated with ibrutinib relapse and fail ibrutinib because of a specific mutation in how that drug is impacting the tumor. But we didn't recruit studies looking just for that. They allowed in other patients who failed ibrutinib for intolerability or other reasons that are not known. What we're seeing now is no difference between those groups and that almost everyone has a strong response post-ibrutinib in a pretreated population. To get something like 85% overall response is impressive to us. And we're doing that with a side effect profile that doesn't have some of the heart and other side effects, the bleeding risks that we see with ibrutinib. So not only may we have a great second-line drug that has at least the potency of the first-line drug but also might be cleaner. That leads us then to the next thing, which is, okay, well, let's go study this in a head-to-head setting, MCL, CLL, both of interest here we announced in December, we're going to do that. That will take a lot of time because the patients do well in this condition. But in the short term, certainly, we'll be looking for a second-line positioning for all-comers because this strength of data demonstrates that that's going to be of high utility, and then through time, perhaps even displacing the first-line therapy.

Yes. So it certainly looks like some great data. It'd be interesting to see how that evolves. Maybe I'll talk about oncology. We're going to be watching Verzenio this year with the adjuvant opportunity. I think we talked a lot about the size of this over time. I think the questions we're still getting is the uptake curve. So how do we think about how quickly you'd envision physicians adopting a data set like this? Can you just talk a little bit about -- do we need to see more data here to convince physicians? Or do you think with the data we have today, there's a compelling argument to bring a lot of these patients on board with the drug in that population you've identified of higher risk adjuvant patients?

Yes. Well, it's not approved in adjuvant setting yet, important to note. We hope that will happen soon. But I can tell you from conversations with thought leaders and the reaction we've seen so far is that the data today is more than sufficient to add to the current therapy regimen of CDK 4/6, and specifically Verzenio, remembering a 25% reduction in risk for patients who relapse after the original disease, there's no cure. So preventing that is a high priority for breast cancer patients and for their doctors. So this is, I think, a breakthrough set of data. The market, as we've talked about, is about 40% of the size of the metastatic opportunity. And it looks like it will be some time before this data could be replicated by another medicine. So we've got a great opportunity once approved in the setting. I'll also remind investors that, by the time we get approved, there'll be more data, right? So we continue to follow all these folks, and the data will mature. And the key marker of invasive disease is strong even at the early time point, and that's historically been quite predictive of things like survival. So that's -- the data will mature. We're confident in the data set. We think there's enough now. I'm sure some experts are even using the drug now off-label. But once approved, we expect a very strong uptake in that setting. I would also point out that I think we are seeing, just based on the totality of information about Verzenio, including OS benefit in the MONARCH 2 population, the only drug with a third-line indication, that moods are changing a bit. We saw new starts in the metastatic market in the U.S. basically double last year, which was an acceleration from the prior year, exiting over 1/3 of the market going to Verzenio. So I think that's another convincing thing, is that the totality of evidence is the strongest for Verzenio versus other CDK 4/6 inhibitors. The doctors are taking note of that as well.

Yes. Absolutely. And then as we look forward to '21, I guess one of the kind of readouts we're watching here is mirikizumab in UC. Can you just talk a little bit about the -- and then more broadly, your immunology portfolio here. How do we think about this business beyond Taltz? What are you most excited about? And how do you see this kind of evolving within Lilly with again, I guess, a couple of different readouts as we go through this year.

Yes. I think there's 3 big ones. We have the mirikizumab data, and that's an important IL-23 inhibitor, that has demonstrated good efficacy in psoriasis. But from the beginning, our interest was really, given the relative saturation of the psoriasis market, products like Taltz, which clear -- 80% of the patients almost completely clear. There's not much headroom left for the dosing regimen in other convenience plays. And Taltz continues to perform well. So UC and Crohn's are really the third and fourth biggest segments for chronic anti-inflammatories in the immunology space. And we see IL-23 is just playing a very significant role there. So we expedited that program. I think we'll be first to launch in ulcerative colitis; Crohn's, likely second. I know risankizumab announced some data just recently. That's strong as is our Phase II data there. And so I think IL-23s are coming, and products like Remicade and Humira likely will be displaced through time by this new or better mechanism. We've got to compete. Immunology is crowded, and it's very competitive. But that data is coming out. We look forward to submitting it and getting going with mirikizumab in UC. And these patients are miserable. I mean it's a tough condition where standard of care Remicade or anti-TNF inhibitors are giving people remission rates in the high teens. That's really poor. And IL-23s promised to at least double that. And I think that's good news. Mirikizumab, we think, is well positioned there. Lebrikizumab, which was last year's JPMorgan conference [indiscernible], Humira is in Phase III. We will get that data this year. So that'll be exciting to see in atopic disease, and there's a lot of discussion from the competitors about the size of that market and the potential there is large, although less developed than some of these other immunology markets. Perhaps competition can help develop that further. And we like some of the features of lebrikizumab on dosing and maybe on side effects that could be differentiating in that class. And then finally, Olumiant or baricitinib has 2 important events this year. One is the maturation of the SLE or lupus data as well as the alopecia areata, which is another derm indication, a little more niche-y but high unmet need in a group of patients, and we've got some profound data there. Meanwhile, we continue to execute against the derm and room opportunities we have for both Olumiant and Taltz. And then we have a strong pipeline in early phase immunology focused on kind of beyond cytokine targets. Immune resolution, which is a new field, but the idea here being, rather than reduce the amount of kind of gas being put on the fire, apply more brake, the body's own natural system to counter-regulate overimmune response. Can we amp that up? And is that the defect in some of the patients? And can we counter-regulate the immune response which may promise resolution for some of these conditions? So that's an interesting portfolio coming and something we'll be talking about more as the next year or 2 progress and those proof of concepts read out.

Excellent. Well, I think we're just about out of time. A lot of different topics we'd go if we had more time, but congrats in all the progress in the business, and I appreciate the time today.

Thanks, Chris. Appreciate it.

Yes. Have a good day.