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

Hi, everybody, and thank you for joining us today. This being the Operational Decisions Conference, we are trying to get to some deep topics, not cover everything but cover a few deep topics. And Lilly's been kind enough to share with us today not 1 but more of the 3 kind of like senior managers with specific topical responsibilities. So we're going to have to go over the next 45 minutes relatively quickly through discussion with those 3 experts. We have Mark Mintun, which is the Head of R&D in Neurology and Pain and the President of Avid Radiopharm, so he's the man on everything Alzheimer's. We have Jake Van Naarden, who is responsible for Oncology, I guess soups-to-nuts, all the way from R&D to commercial business; and Jamie Croaning, which is responsible for the tirzepatide program. Thank you very much for all 3 of you for joining us today.

Thanks for having us.

Delighted.

Thank you. So we're going to start with Mark, who has an obligation at 1, and we'll start with a question for him. And we'll try to kind of hit it 15 minutes or a little bit more for a member of the team, starting with Alzheimer's. So Mark, one of the things I really enjoy reading on clinicaltrials.gov is your plan for the prevention trial for donanemab. Can you just go through this trial in a bit of details explaining what you're trying to achieve, how you're selecting the patients and the outcomes that you're -- the outcome measures you're using?

Yes. Yes, sure. So it's a trial we're really excited about. Clearly, we have already a hypothesis that if the disease is very late in the stage and the tauopathy has spread, it's going to be very, very hard to stop it. Of course, the opposite concept is that the earlier we are, perhaps the more dramatic percent slowing, even maybe 100% stopping of the disease. So we look at the amyloid as inciting and accelerating Alzheimer's disease. So if we go to patients who are essentially asymptomatic, so we're going to be enrolling patients from 55- to 80-year-old that are asymptomatic there. We're going to be starting them. They're going to have -- we're going to use as -- a clinical dementia rating global scale of 0, which means that there's no evidence of cognitive impairment or dementia, and we're going to enroll them. So the exciting concept is that we're going to get these patients to be symptomatic. But the problem -- it's a -- I've been working in this area for a long time. Lots of people say they're going to do that, but it's how do you do it. One of the ways -- the other exciting thing is we're going to use a blood test, our P-tau217 that has performed so well and has been published in different reports, as a way of detecting the presence of Alzheimer's pathology in asymptomatic people. So we're going to use a blood test to demonstrate that they have the Alzheimer's pathology. And then we're going to have a short course of donanemab to remove the amyloid part of that pathology, which we hope essentially sets them back -- after 10, 15 years of maybe accumulating amyloid, we're going to hopefully put their disease back all the way back to the beginning as we treat them for that. And then we're going to use an event-driven concept, watching them, we think at least 3 years it's probably going to take, as we watch for those people who convert. So half the people will be on placebo, half the people who have been -- received donanemab treatment. And our hypothesis is that the people who do not -- when on the placebo arm are going to convert into having some -- a positive clinical dementia rating at a much higher rate than those who have been treated with donanemab. And then the final extra little kicker here is that we're going to do this in a decentralized way. So we're going to be having a different modus for being able to enroll these patients. Because of the reliance on a blood test rather than some exact cognitive score, we believe we can enroll this much faster and be able to get to an answer much quicker. So it incorporates a lot of, we hope, novel and successful ideas to be able to bring this type of technology to a patient group that usually had to wait. People with amyloid disease, they have to wait until they actually have a substantial amount of symptoms and therefore potential brain damage. And we hope to stop this disease before that.

Okay. So no PET scans and actually no -- not a detection of amyloid at all. You are detecting them with tau.

Yes. So we're going to...

Beginning of the -- the beginning of the cell death, the beginning of the damage in the brain. And you're going to be able to do that, do essentially a decentralized or, shall we call it, online video cognitive test that you're going to do this. And these are your positive markers to define somebody has Alzheimer's, okay? That's actually a new definition for the way we select patients for those trials, one which is much more interesting to use in the commercial setting, right?

Yes.

Because what's delaying the use of a lot of the Alzheimer's drug is just going through a PET scan, going through other tests before you can get enrolled. So you're going to have the tests available more publicly. So if I want tomorrow to go out there and get tested for having Alzheimer's without going through a PET scan, I can't do it today. When is this test coming to the marketplace? And when are you going to get it validated by FDA so we can start having blood -- can I say broadly -- I need that to have an Alzheimer's medication apparently. So broadly based testing of the public for the presence of Alzheimer's?

Well, we don't want to get too ahead of ourselves. When we use this test in a clinical trial, we can accept certain types of sensitivity issues in this particular trial population. So as we continue to refine this test and look at the ways of scaling it, we are pretty comfortable that before donanemab launches, we'll be able to work towards getting this probably with a commercial partner, but we'll be able to work towards getting this in the form of a laboratory-developed test for patients to use. I think a fair amount of the data will support the concept that it will potentially reduce the need of PET scans. And so in the end, I don't think this will make PET scans like completely unuseful.

Sure.

But I do believe that for a sizable percentage of patients, this will be able to give -- we hope if everything goes well, this will be able to give a pretty definitive answer. And so we're working towards -- on that. Before the end of '22, we hope to be able to have this out there and preferably before the beginning of -- if donanemab is successfully approved for accelerated approval before the -- before we actually launch that. So we have those types of targets in mind for being able to get this so that people can actually benefit from this type of diagnostic information.

Could you also use it to follow the progress of the disease? Essentially, one of the things you've pioneered is this idea, well, we only treat until we are done with the plaque. This doesn't really tell you much about the plaque, but you can detect a reduction in tau in the blood after treatment. So how do you think about that as a marker for ongoing treatment or as a way to follow the patient course to make sure the patient is benefiting there?

Yes. We -- as we've shown -- we've already shown in our TRAILBLAZER-ALZ study that this P-tau217 definitely drops with onset of treatment with donanemab.

Yes.

We've not shown -- and obviously, the data is still young, and this is a test that just burst on the scene just a little over a year ago. So longitudinal data is not what we're good at right now. We don't have a ton of longitudinal treatment data with this marker. I'm not confident that this marker will give us ability to tell us that the amyloid is removed from the brain. At this moment, I would say that we're going to be stuck with PET scans to show that someone has been reduced to no amyloid. I think the field is moving rapidly. So of course, we're going to be working with the people who are pioneering different markers in this, and so I'm not saying that, that won't be something coming up in the future. But at this moment, P-tau217 is reporting on a type of pathology that occurs with -- a type of tau pathology that occurs with Alzheimer's disease. And unfortunately, you can remove all the amyloid in the brain, or at least all the visible amyloid on a PET scan, and you still have tau pathology in the brain. So we are not expecting the P-tau217 to go back to normal. We're really glad that it reduces dramatically, but that's something we hope about the leading edge of the tauopathy. But I don't see the tauopathy going completely quiet in a person with established Alzheimer's disease when that means it probably is not going to be a one-to-one predictor of amyloid.

Or more correctly, a drug will actually not completely stop Alzheimer's progression. We'll probably need other tools as well to get there. Okay.

That is a secondary, yes, conclusion, exactly. Exactly.

So you are seeking an early approval. And it seem from the clinical -- from the -- what we're seeing in the market today that physicians really want proof of clinical efficacy, not biomarkers, before they wholesale buy into this, the concept that reducing amyloid helps with [indiscernible]. So it's understandable, and your initial view is understandable as well. Roche has taken the opposite road. Roche has basically said, look, we're not even going to bother commercializing this product before we actually have proof of clinical benefit. And the question is, given what we're seeing with -- commercially today, why not -- why shouldn't Lilly follow the same approach? Why not wait until you have clinical benefit and only then launch the drug?

Well, it's an interesting question. And of course, it is interesting, the difficulties that are facing aducanumab's uptake in the market. Now I would argue that we're a little different. I don't think there's a direct comparison. And we're looking at a situation where we would be potentially coming under the market in the second half of '22. At that point, there could be an NCD, which is the national coverage determination by Medicare, that might clear up a little bit of the uncertainty right now. And then the other aspect is the combination of what we see as the demonstrated biomarker, really the amyloid -- demonstrating a much higher amount of amyloid. And of course, we're going to be testing that hypothesis in a head-to-head. And then the other aspect is that we actually had a trial that we ran, admittedly a Phase II-sized trial, but if we trial -- that we specified the outcome. We ran it all the way to the end. It was a clean, it was positive and it's associated with multiple biomarkers. So the data for each drug is not identical, and I think that people will see that there are indeed advantages to our overall package. But even that said, we have room to saying there could be tremendous headwinds to -- and in a way, we just want to embrace that. We know -- we want to be able to get the drug because patients are going to be able to benefit. That's our premise. We need to build the diagnostic ecosystem. We need to understand exactly the right way that a P-tau217 assay might be used or maybe an A-beta42/40 assay. Or where does the PET scans play the most critical role? Do we have the infrastructure for blood testing in -- on PET scans? And then, of course, the normal logistics of ensuring -- starting that conversation with reimbursement and infusion capacity. So those things -- the earlier you start that, I think the earlier you are able to iron out the difficulties. And we just sort of feel like that we just have an impatience to get on with this.

I got it. So it's a little bit more like, look, we -- it will take us some time to get it right. If we're in the market a year before we actually get our clinical results, we have a chance to make sure that things work.

Yes. Our confirmatory clinical result. Our -- so confirmatory clinical result.

I stand corrected. Phase III confirmatory pivotal. I do hear the focus on the NCD, but it always a bit confused me. If the trials that come out between kind of like mid-'22, to mid-'23 come out negative and people are saying the NCD is critical, if they come out negative, then the NCD would not make a difference because the FDA will take those drugs on the market because the hypothesis that the biomarker is right is just wrong. And on the other hand, if the trials come out positive, then whatever the NCD is, presumably, the CMS will change their view to give broader access to these drugs even before they're approved. So I never quite understood why the NCD is important. It seems it's -- it all should be about the clinical outcome. And I'm not quite sure why the focus is on the NCD. I mean maybe you can explain to me from a pharma perspective why that is the case. Because I've seen a lot of discussion on this, I just never understood the logic.

You asked if I could explain it, and I would say I'm not 100% able to explain the logic there. I think there are some holes in the logic. I think that there -- this is perhaps the consequence of the field being -- moving in slow then fast and then slow.

Got it.

It is a transformation of the whole thing. We simultaneously are talking about new ways of diagnosing and new ways of treating at the same time of something that has been considered untreatable, and that's been baked in the way doctors and patients have been thinking about this for 20, 30 years. So I guess I'm a little bit sympathetic that this is going to be a bit of a bumpy ride, that these type of major changes don't happen overnight and they don't -- not all the different stakeholders in this are going to move in lockstep. But I'm very sympathetic with your -- the description of the problem you just pointed out. And I would imagine that if the Medicare -- CMS asked me sort of my opinion, I would say build into any decision you make the concept that -- of both of those exact outcomes that you just talked about. And I think that just needs to be part of the equation. And perhaps even do that on a drug-by-drug basis because one drug could fail a trial for reasons that everyone will look at it and say, well, there was something difficult about the way that trial ended up having to be run. That's certainly not new in Alzheimer's disease. And then another drug could fail because it actually is demonstrating no efficacy. So we do have to keep that in mind as well.

Okay. We've got about 5 minutes to talk to you, Mark, so let me hit a couple of 2 or 3 points really -- real quickly. One of them is this idea of BAN2401 coming in with the proposition of similar efficacy but very low -- oh, very low ARIA rates because of targeting the oligomers as the main specie as opposed to the plaque and yet indirectly removing after 1.5 years roughly the same amount of plaque as the direct plaque removers. How do you think about this drug? What are the gaps here? Is this really the drug that you consider to be the main drugs you need to be concerned about as opposed to Aduhelm? How are you thinking about that one?

Well, some of what you said, I guess, I'd take issue. I know that lecanemab, BAN2401, is measured in a preclinical setting of binding oligomers, and it also binds the plaques. And I would argue that the part that binds the plaques and initiates a removal of plaques probably through microglia recognition of that binding is, in my mind, very likely the source of the efficacy. So it may or may not bind to oligomers that you can prove exist in a test tube. It is much, much harder to show that, that's adding additional efficacy in some way. And I would argue that our drug, which we have no evidence binds to anything other than insoluble plaque, no -- nothing floating around or anything else can be detected to do that. And I would not -- our efficacy matches or exceeds the efficacy of those drugs that say they bind to 2 things and therefore might give 2 different ways. I would say that is an argument for saying there's not much additional gain there by doing your -- getting 2-for-1 type of binding. I think it is really important that -- to comment on safety, though, moving to your comment about ARIA. ARIA itself is an MR finding. The important concept is sort of symptomatic, things that actually affect people's lives, things that cause stopping of doses and things like that. And of course, I think the real test is going to be our Phase III studies. They -- it's really important that they complete their Phase III study for the field. And we see exactly what the risk-benefit ratio is there, and it's obviously really important for our confirmatory study with a much larger number of subjects and including subject with both, as we put it, intermediate and high tau, the whole spectrum there. So I would say we sort of are going to have to wait until those types of Phase III studies report out so that we'd be able to sit there and look at that type of safety and efficacy data.

Okay. And another development in the clinic that I kind of noticed about a year ago is the effort by Roche to have a shuttle that moves the protein into the brain, which seems very elegant. I think the early data showed that it can move the brain partition from 1 in 10 to like 1 in 1.5 or something of that nature, blood versus brain. How are you guys thinking about this? Is there development in this field? Is this something you guys are also working on? Stay tuned or I guess something we're going to show you soon? Or is this still a science that is yet to emerge?

I'm trying to figure out which -- a multiple choice question, huh?

Yes.

So the first [indiscernible] exciting. And I applaud them for going forward with that. That's really exciting. I also -- and we are watching that area. We are indeed active and thinking hard about the ways we can contribute in that area with a variety of different things. I would say the one thing that just you and the audience seems to be very careful about is that one of the odd things about binding a transferrin receptor is that it clears your drug really rapidly. Not always, but there is a phenomenon where you end up clearing it. So it is actually not available to the organ of interest, in this case the brain, as much. But -- and so you can imagine a situation where the exact same amount might be getting into the brain. But with a much faster clearance, you now are left -- the brain is up here, the blood is going up here. And you say, wow, we now have like a tenfold increase in the brain-to-plasma ratio. The most important thing is to say -- and I'm not being -- this is not new. I'm not revealing some huge thing. It's just a caveat that one has to remember when one looks at these things, is are you getting in an increased amount of functional drug and that the PD, the pharmacodynamic measure, is actually increased per gram of drug you're giving? And -- but do I think it's possible? Absolutely. I think there's progress being made at multiple different places, and I really applaud Roche for going forward and getting their data out there. And we're definitely excited about this area. I would say I'm not 100% certain I need to be excited about it for donanemab. We're getting great plaque clearance. I don't think I need to go in there and try to get more. Or -- I'm not trying to mess with that at the moment. So I don't think there's an upgrade to donanemab on my future in that regard. But I do think that from a field of overall delivery of drugs to the brain, it's a huge, important area.

Got it. Last question. The safety addendum to TRAILBLAZER-2 that was added, there were some questions about that. Can you explain what is the objective here? And what does it do?

Yes. I think some of the questions that I've heard are sort of like, is this something the FDA asked us to do? And I can put that to rest. There was no request by the FDA to do this. This is an addendum that would dramatically increase the number of safety exposures. Obviously, we finished TRAILBLAZER-ALZ with a small Phase II. And so it is great that we're able to recruit at a very high rate, by the way, this addendum to increase the number of safety exposures. So these are open label, people who are going on an open-label trial. And this additional safety data will be part of supporting the rolling submission. So this is something -- we're not doing it because of a safety signal, we're doing it to increase the total amount of safety data we have.

Got it. Mark, really appreciate your time today. I know you're very busy. So thank you for the time and always finding interesting insight when we have a chance to discuss this.

Well, you're very welcome and thank you for the great questions.

Jake, thank you. Let's move over to you just because you're kind of touching -- you're responsible for some of my greatest hobbies, which is looking at the way the U.S. drug market is changing over time in terms of incentive structures. So you are developing a PD-1, a low-priced PD-1, which is supposed to be approved in the first half of next year, a PD-1 from China, Tyvyt, that is supposed to be -- I called it low price, I probably should not -- an excellent quality PD-1 coming from China. And I was wondering if you can talk a little bit about the timing for various indications coming online, the quality of the overall programs and how are you thinking about this drug playing in the market.

Sure. Happy to. So we're talking about a drug called sintilimab, but the brand name is Tyvyt in China that we in-licensed. We've been engaged in a partnership with Innovent in China for a long -- for many years and have economics on the sales of the product in China. We, last year, expanded that collaboration to give Lilly pole position and rights outside of China. And let's talk about the reason we did that. Innovent had done a really nice job running a bunch of randomized trials, but the one that really piqued our interest was in lung cancer really with a dataset that looks very similar to what we have with pembrolizumab and KEYNOTE-189. Obviously, that's the largest driver of sales for the PD-1 class in general. And seeing a dataset run at such a high degree of quality with both active and control arms that look so similar to that which was run in a global study gave us a lot of confidence in the drug and the quality of the package, and really kudos to Innovent for doing that. And so we saw an opportunity, admittedly an opportunistic one, to bring this agent starting to the U.S. and perhaps eventually to Europe and other markets to increase the competitiveness of the space specifically by pricing it at a substantially lower list price than the agents that have been introduced. That BLA was submitted to the agency. We have a PDUFA date in the first half of next year. And I think as folks know, there's an ODAC being scheduled as well. So we'll see if the drug gets approved. That's not a guarantee by any stretch. As I'm sure you and others know, there's been a lot of public commentary about the degree to which clinical datasets enrolled entirely in China are applicable for a U.S. population or approvable for U.S. population. And so we have to see how that goes. As far as the clinical review itself has gone, fine so far. No, we haven't heard anything specific from the agency with respect to the actual package. But I think there's probably some policy wins that are being discussed is my speculation.

So the interesting part is, is this -- do you know that this is the kind of the main question that the ODAC is going to be asking?

I don't.

So you just don't...

We don't have any insight on the topics of the ODAC.

So what you're saying is the issue of the source of the data is your speculation about what this is about? Like...

Well, we know that this has been a topic of both public -- I mean, we know this is a topic. I think the public knows this is a topic. So I don't think there's any surprise there. Whether or not that's like the main focus of an ODAC, I don't know.

Okay. Some of the other drugs coming from China -- I think there was just an announcement by one of your peers that they received an accelerated review with no ODAC. So I kind of agreed with you when you said it on the call, on the call...

Maybe there's an important distinction to make there actually, which is that there's been -- there are probably, broadly speaking, 2 different categories of indications or submissions for drugs like these. One is a sort of finding a niche tumor type that hasn't been explored by existing PD-1 inhibitors and sort of -- and therefore has the flavor of differentiation. We can debate whether that's real or not. But as that flavor, I think that's what you're alluding to in that program versus what we're doing, which is clearly an indication that other PD-1s have where the innovation, so to speak, we're bringing to the table is really about patient affordability.

So you see zero crossroads there? Because I got to agree with you, there's slight -- there's a difference there.

I see very little sort of crossroads or read-through between those 2 -- I think they're just -- they're being treated differently, let's just put it that way.

Okay. So the other thing you kind of -- you mentioned is that you're aiming for a low-WAC model. And the question I always had about this is, look, yes, I understand the low-WAC model. It might be relevant for Kaiser Permanente or IDNs like that. But there's a very large segment of the market, which is physician, oncology practices, that would like a low price but would like you to start at a similar price to the branded and then come down over time as opposed to go with the low WAC and stay there. And I was kind of thinking about -- have you kind of thought about these 2 models? Is there like a statement you're trying to make with a low WAC? Or is this just what you think is going to be the better business longer term?

And we've thought about a lot of different strategies for how to go with this. So I can assure you we've thought about exactly what you're alluding to. Look, CMS has or, I should say, is running a pilot program, the OCM model, explicitly to incent the kind of behavior that we are intending to do here. And so you're right, there are going to be, frankly, somewhat large segments of the market for which this is not going to be appealing despite that statement to -- I think, to a lay audience being a confusing thing to say. It's just true, unfortunately, for the way our system is structured. We think this is the most effective strategy to lower patient exposure, which is really the goal here, and system exposure, so to speak. And there are a handful of very high-quality practices that have bought into the OCM model and I think will be good counterparties and will find this attractive.

Yes.

But in fairness, like is this going to be a material product for Lilly? I doubt it. I mean this is a niche idea, one that we saw a very specific opportunity to do in light of a dataset that already existed. So I think within the Lilly Oncology franchise more broadly, we have more interesting things going on, frankly. But I understand why from an industry perspective, for folks like Merck and EQRx and others, this is an interesting topic. But it's not one that consumes a huge amount of our bandwidth, to be frank.

So to be clear -- so perfectly color. I got half a dozen more question on this one, but you kind of made a good point. So why don't we move to a -- more to move of your -- some of your other interesting drugs in development?

Sure.

One that you are pursuing -- one opportunity that's kind of interesting is the BCL-2. And you're only the second company to pursue a BCL-2. So can you first discuss the logic of why you are developing one when there is a perfectly good BCL-2 in the market and nobody else seems to think there's a need for another BCL-2? What is your logic for developing this drug?

Funny, this is a similar extension of the prior conversation. We -- a really important franchise product for us is pirtobrutinib, the BTK inhibitor that we think has a meaningful role in the treatment of CLL and mantle cell lymphoma. And particularly in CLL, they are basically just too abstract. There are basically 2 modern treatment paradigms for a BTK inhibitor in the disease. One is as monotherapy given continuously until progression or intolerance. And the other is given in a time-limited fashion in combination today with venetoclax. And unfortunately for patients today, there are no companies that have both a BTK and a BCL-2 inhibitor in their own portfolio. And so despite that time-limited regimen being very interesting for a variety of patient settings, you're talking about combining 2 extremely high-priced agents. And so we just see an opportunity to be more competitive and do a better job by patients by owning both.

That's just interesting. So I didn't even realize you're thinking about that one. So you are thinking about this combination of BCL-2 and BTK inhibitor essentially, at least partially to be driven by cost -- overall system cost consideration as opposed to making a better BCL-2. Can this...

Yes, we don't expect our BCL-2 inhibitor to be clinically differentiated from venetoclax.

Interesting. So one other thing I was always wondering about BCL-2 is the people using small molecules here. And the question is -- BCL-2, what they do is they target a membrane -- they target BCL-2, which is a membrane coating on the mitochondria. And they cause a lot of the side effects because there's a rapid action that kills those cells. I was always wondering why people are not targeting degrading that protein or using antisense, preventing its formation as a way of slowing down that death of cells and thus getting rid of some of the side effects.

Well, the side effect you're referring to is really tumor lysis, I assume?

Correct.

I mean outside of TLS, venetoclax actually is incredibly well tolerated. Yes, I mean, it's an interesting idea you're invoking. I don't know if folks are pursuing that. I think that as you probably remember from way back when, the venetoclax development program took a lot of twists and turns. The TLS signal sort of came out of nowhere, and then there was a clinical hold, and then they sort of had the sort of backtrack and think again about how to dose it with the dose ramp. So there's -- we've landed at a place with venetoclax that maybe wasn't super linear. And with all the right dose optimization done with that agent from the beginning, probably not just because of the path that it took. Whether we can differentiate there, maybe we can, I don't know. Obviously, we have the benefit of learning a lot from that agent and that program.

Yes, I guess. So this comes back to this issue of BCL-2 was really thought about not as can you get a better BCL-2, but there was a need for a second BCL-2 in the market for commercial reasons, and that will be reason...

I just think -- if you take a step back, I think even if you look at analogs outside of oncology, if you know there's an important regimen in your portfolio that relies upon a combination, it's generally better if you can own both parts.

Completely agree. Okay. So let's go to the agents that might be differentiated, which is a BTK inhibitor. And of course, less about your ability to differentiate versus ibrutinib and more your ability to differentiate against the second-generation drugs, Calquence and Brusinka (sic) [ Brukinsa ]. So can you just describe how you see your molecule differentiate from those second-generation drugs?

Sure. Yes. So our molecule is pretty different from those drugs both in its mechanism and frankly its clinical data. So ibrutinib, acalabrutinib and zanubrutinib are all covalent BTK inhibitors. They bind in the same way. They have different selectivity profiles and, as a result, have marginally different tolerance profiles. And I would -- I think acalabrutinib and zanubrutinib are both probably marginally better tolerated than ibrutinib. They are no different on efficacy despite some of the more recent claims from some of those agents. They're equally efficacious as -- all 3 of those agents. Now you can see in the script data that the tolerability advantage of a drug like acalabrutinib is playing out. It's taking share. Okay, fine. That makes sense to me. It's been largely an elderly population. And if you can make no compromise on efficacy and improved tolerability, why not, right, at the same price point or a similar price point. So I get that. But what's important is that these drugs do not work in patients who progressed on any one of them. And unfortunately, that's what happens. And so while -- you can cycle through these drugs in the context of intolerance. But in the context of true progression, which is an unfortunate, inevitable reality for all these patients, you can't cycle through them. And because of the way pirto works, it's a reversible inhibitor that binds differently and has very different pharmacology, we're able to induce like bona fide durable responses with a substantial clinical signal in patients who've actually progressed frankly on all 3 of these drugs. So the initial...

So...

Oh, sorry.

Go ahead. So I was going to ask a question on head-to-head. I think you're doing a head-to-head trial already against ibrutinib.

Yes.

Are you going to do it against acala as well? Or is that just not necessary?

I don't think -- personally, I don't think it's necessary. Again, our head-to-head ideas are about efficacy, not safety. So when you look at the head-to-head studies that acala and zanu have run, those are safety head-to-head experiments. And frankly, the efficacy data, you can see the curves basically overlap. I think if you're going to run an efficacy experiment, I'd rather run the experiment against the market leader with ibrutinib. And again, I don't think, on the efficacy side, it really matters. They're all sort of the same.

Got it. And the time lines for this? Remind us.

Oh. So the head-to-head study against ibrutinib in mantle cell is a study that's going on right now. We just started earlier this year. The head-to-head in CLL against ibrutinib will start in the beginning of next year. Both of those are long studies. I mean that's -- they're both going to take a while. But I think it's important to say, and I was going to say this earlier, the initial idea for this drug, pirto, is really in patients who've already been treated with these BTK inhibitors. So the head-to-heads are to bring these drugs into an earlier line of therapy. But frankly, in today's treatment paradigm, the real unmet need is in patients who've relapsed on these drugs because the options there are limited. You basically have venetoclax, which is very effective but is complicated to administer for the reasons we were alluding to earlier. You have PI3-kinase inhibitors, which are toxic and don't work that well. And then I guess you can re-treat with chemoimmunotherapy. You can try CAR-T. There's just not -- there's very, very few reasonable options for patients. So to have -- to be able to essentially continue your pathway, so to speak, so from a covalent BTK to a reversible, well-tolerated oral medicine, we think that's a very attractive option for patients. I think our -- the investigators who've worked with the drug agree with that, and practitioners, more broadly, have seen the data. Obviously, we have to get the drug approved, which is its own challenge, but the initial home we see is really in BTK-treated patients.

Okay. But you're not giving up on the first line just because -- obviously, it's a much bigger -- commercially a much bigger target than being on second line.

Well, we should debate that a little bit. We are not -- you're right, we're not giving up on first line. We're starting a first-line study against chemoimmunotherapy literally right now. So no, we're not. But the reason these drugs are commercially interesting is because of duration. And we think that we have the opportunity for a very long duration of therapy and disease control in the second- or third-line setting. So -- and again, this -- it's not as if CLL is a common disease.

I guess most patients just typically go on some sort of a BTK inhibitor and a second drug early on. And the second lines are there, but first-line remission could be 3 to 5 years, second line is going to be 2, right? So I think your point is...

Well, we have to see how our data play out.

Yes, exactly. You can probably go out longer. Okay. You mentioned the PI3K3A. I was wondering if you can talk about it a little bit more. I think that's an asset most people know less.

Yes. This is something we just started talking about. It'll be going into the clinic next year. This is a really exciting, new program for us. So it's a -- well, I'm going to describe it. It's a little bit of a mouthful, but -- so we can unpack it. It's a mutant-selective PI3-kinase alpha inhibitor against a mutation called H1047R. And this is a mutation in PI3-kinase alpha that accounts for about 15% of breast cancers. It's clearly a driving -- driver mutation. We, of course, have the approved drug, alpelisib, with randomized PFS benefit in this population. However, the drug comes with a lot of toxicity because it's a wild-type inhibitor, and it inhibits wild type and mutant to the same degree. Inhibiting wild-type PI3-kinase alpha is not a good thing for the healthy cells in your body. And so you see a lot of insulin derangements, fatigue, GI distress. There's a lot of side effects that come along with that. And so our idea was, well, is there a way to bind the mutant form and spare the wild-type form? And that's what our drug is able to do. It really doesn't inhibit wild type at all. And so we're able to induce preclinically meaningful tumor regressions in these H1047R-mutated models with literally no impact on insulin or C-reactive protein at all. It's a really interesting, cool molecule that we'll put into the clinic next year. And obviously, with the breast cancer franchise that we have with Verzenio and imlunestrant and SERD, this is just another important advance for us for men and women with breast cancer.

Are women typically screened for this mutation on a regular basis? Or is this something you hope to introduce?

The answer is sort of yes and no. So obviously, the introduction of alpelisib has increased screening for this mutation but not to high levels, frankly, because alpelisib is such a toxic agent that it hasn't "catalyzed" that activity to a great degree. That having been said, testing for this is fairly simple. And I think with a drug that has more activity and is better tolerated, that's not going to be a big hill to climb, to be honest. And it's so common. 15% is -- it's like EGFR in lung cancer.

Right. It's -- and that's a pretty big drug. So I got to finish with KRAS. We got a couple more. We got to keep KRAS. So there are 2 questions here. One of them is, you kind of took a very strong position in this new poster, saying, "I have a better KRAS," and I want you to talk about that. And then the second point is there is like a concerning data point that came out all the way from AACR this year, which basically -- or the observation has been that a lot of those tumors, at least in second line, throw out a lot of ways to overcome the KRAS block, a significant KRAS block. It's both an -- either overactivating KRAS -- the other, throwing a mutation that offsets the drug, amplifying out the path of the tumor. I guess the question for you is, in your mind, how clear is it that we can find a combination for KRAS that will allow us to have -- to kind of block the, MAP/RAS (sic) [ RAS/RAF/MAPK ] pathway and thus be able to get good efficacy from this? Or is there like a real concern that this will remain active but not very long-duration data? Essentially, those are the 2...

I think you're asking the right question. So let me just start. Number one, we've done some preclinical modeling around target coverage, and we think we have the ability to inhibit the target more deeply than the 2 other agents. I'm not ready to say we have a better KRAS inhibitor by any stretch. So if that was the implication, I -- we should back that down. We're in dose escalation. We'll see what we have. We'll see if our human target coverage stands up to that. I think that in many other target classes, more target inhibition is better. So if that turns out to be true here, then maybe we can see more substantial monotherapy activity versus what's been observed with sotorasib and adagrasib. We'll see. I think the second component is about combination therapy, which is a crucial question because if this entire class ends up truly being relegated to monotherapy in second-line lung cancer, it's just not, unfortunately, that important for companies or for patients. It's a nice advance for the field, but it's sort of a modest treatment effect size in that setting. And so I think the folks, both clinicians, patients and companies, are all trying to figure out, well, how did these drugs get into the first-line setting of KRAS mutant lung cancer. And that's where the important combinations come in, whether it's with PD-1 inhibitors, chemotherapy. Both were sort of novel, novel combinations. I think just reacting to what we've all seen publicly, we've not seen a lot to make us all bullish about that just yet. And whether that's drug-drug interactions that have been a problem for some of these agents or other things, I don't fully know. It's an opportunity for us to explore with our drug that could our drug be more combinable than some of these other agents? I don't know the answer to that, but it's something we'll be looking for in the context of the expansion cohort of the study. The last point you made is about resistance. You're reacting to the Ryan Corcoran paper, I think, in cancer discovery earlier this year. It's a very small sample size. I don't know how to interpret those data, to be honest. I mean this is a pretty common variant in lung cancer, all things considered. Let's see a big dataset of how resistance emerges before we start getting into that. Interestingly, though, these covalent G12C inhibitors, let's call it ours, Mirati's, Amgen's, they actually all have slightly different mutational coverage within those variants you alluded to. So whether that plays out clinically, I have no idea. But it's in vitro. They're not actually the same in this respect.

Okay. Very good. Jake, really appreciate your time today.

Yes. Thank you.

So Jamie, thank you for sticking around. And it took a couple of minutes more than I thought. There are a couple of points here I wanted to do. It's kind of funny because you actually have probably the biggest near-term catalysts for the company, your portfolio here. But I guess I kind of want to touch with you on 2 or 3 things. First of all, on GIP/GLP, then about obesity, then a little bit about HFpEF.

Okay.

So starting with GIP/GLP. I guess there was quite a bit of discussion among you and your peers about whether adding a mechanism to GLP-1 actually adds anything or we really just have kind of like a well-titrated, well-designed GLP-1, and the GIP thing is nice but really doesn't do anything. So can you just talk a little bit about where you guys -- now that you've got the full datasets, how do you guys think about this issue? What does the GIP mechanism add? And how are you going to use it?

Yes. Well, it's A great question and one that we've, for a period of time here, have been looking at it from an MOA perspective in trying to understand the dual-agonist approach and what the incremental value that the GIP component provides. We've done some early clinical studies to address the beta cell function and adipose tissue metabolism that takes place. But the one study that we just completed and actually was released at EASD was our clamp study that showed -- a 28-week study comparing 15 milligrams of tirzepatide to semaglutide. And we took clamp measurements at different time points as well as a mixed-meal tolerance test to see if we can highlight the improvement in insulin secretion, insulin sensitivity and reduced glucagon. And so those studies read out and now -- and we did present those data. And what we saw was that we did see an improvement in insulin sensitivity in beta cell health compared to semaglutide. And that was one of several studies that we're conducting from an MRI perspective that will help highlight and help us with the understanding of the GIP component. I'd also point to the semaglutide -- or, excuse me, the SURPASS-2 study, where we compared to semaglutide 1.0. In that study, we saw improvements in the lipid profiles of tirzepatide compared to semaglutide, just another added benefit of what we feel the incremental value of the GIP and the GLP component add to the GLP alone.

Got it. And there was also some discussion earlier about antiemetic effect of the GIP mechanism. What's the thinking -- what's the kind of the updated thinking around that?

Sorry, what effect?

The antiemetic, patient has less nausea.

Antiemetic. Yes. So there are some preclinical studies and where we see the impact on the GIP component. What is nice, if you see in our overall SURPASS program, the GI events and the tolerability was improved from what we saw in the Phase II study. From a dose escalation perspective, we went with more of a slow, methodical dose escalation in 2.5-milligram increments. And that allowed us to titrate these patients to their doses. And through that, we saw an improvement in the overall tolerability of the drug.

Anything else, by the way, you will highlight at EASD that we should look at?

Yes. So the important data that came out of EASD was very exciting. As we presented all this data at ADA, we had a very busy EASD as well. So 2 important studies that read out. We had the SURPASS-4 study, where we had top line results back in May where we showed the A1c reduction of 2.6% at 52 weeks. What we were able to provide at the EASD is that, that was maintained up to 104 weeks. So you saw the durability of the A1c reduction out to 104 weeks. And similar where -- in the weight loss where we saw a 13% decrease in weight at 52 weeks. That also was maintained out to 104 weeks as well. So while these data aren't definitive answering the question whether it's long-term durability, it does provide us some encouraging results that we've seen in these studies. And then the second one that I would highlight is we had a SURPASS-3. We had 2 sub-studies that we completed. We did a CGM study where we demonstrated the variability of blood glucose control levels. It was a comparison study where we compared it degludec. And what we were able to see is that the timing range was significantly improved for tirzepatide compared to degludec. And we also did an MRI addendum where we looked at liver fat content as well as adipose tissue. And there, again, we saw some improvement in tirzepatide versus degludec in the liver fat content, 47% for tirzepatide to 11% for degludec.

Oh, GLP-1, a drug we should all take. So you got to ask him for a side question on NASH. So help me out here, a little bit about NASH, a little bit about Alzheimer's. Where do you see a role for GLP-1? I know it's a little bit off the bat here. But since you brought it up, I kind of thought I should ask.

Yes. Well, we are conducting a Phase II study for NASH, where our SYNERGY-NASH study is ongoing. It's in the enrollment phase right now. We do believe that we'll complete that study next year. And we're going to start to readout of that study in 2023. We're excited about this. The data that we've generated with the MRI addendum allows us to believe that there's a -- the weight loss that we're seeing is having an improvement in the overall liver fat content. And we're looking forward to the results from that study reading out in 2023. From an Alzheimer's perspective, we have not initiated or announced the study being completed in Alzheimer. It is an area that we continue to explore as an opportunity for where tirzepatide can provide additional data to help support that disease state.

Got it. So let's talk a little bit about obesity. Kind of can you remind us a little bit of your time lines? And where do you differentiate from our friends in Denmark?

Yes. So we're really excited about our SURMOUNT program. So the SURMOUNT program includes 4 registration studies for the obesity indication. The first one, SURMOUNT-1, is a weight management study in patients with obesity. SURMOUNT-2 is a weight management study in patients with obesity and type 2 diabetes. SURMOUNT-3 is our weight management program after intensive lifestyle intervention. So these patients go under intensive intervention to lose the weight and then they're randomized to tirzepatide to see what additional weight loss they can have. And then the final study is our weight maintenance program, where, again, we randomize the patients, we give them intense maximum tolerated doses of tirzepatide and then randomize them to either placebo or to the active arm and see if that sustainability of their weight maintenance continues out through the period of studies. Our program -- the first results will come from SURMOUNT-1. As we mentioned on the previous call, we anticipate those results to come out in the first half of 2022, and the readouts from SURMOUNT-2 through 4 should occur in 2023.

Okay. Cardiovascular outcome study?

Yes. So our SURPASS-CVOT study is ongoing. So that is our study which is the head-to-head against dulaglutide. That is actually enrolling very well, and we anticipate having readout of that study in 2024.

So that was -- is not part of the registration package but part of the discussion about how do you use those drugs afterwards kind of thing?

So the tirzepatide -- the SURPASS-CVOT is in type 2 patients. So that is an indication that we'll move forward to for an outcome study there. We have not announced a study in the SURMOUNT program -- an outcome study in the SURMOUNT program at this time.

I hope you're not excluding one from coming eventually.

No. We -- I mean, we continue to believe that weight loss with tirzepatide will have meaningful effects, both metabolically and non-metabolically, on health improvements, and we continue to explore the opportunities for us to continue to find opportunities within this space.

Got it. Can you -- do you have a feel for what would it take to change CMS' view about including obesity as part of the things that the federal government pays for? I mean have you -- I'm sure you had discussions multiple times, but do they have enough of a clarity in their mind what do they need to see in order to do that? Or are we not there yet?

Well, I'm not the expert, but I'll tell you my thoughts on this. I do think, as we got into this field, we recognize that continuing education needs to occur that obesity is a chronic disease, and it leads to medical complications that need to be addressed. And so our ability to educate physician, patients, payers of that is important for us to be able to provide these drugs to patients who suffer from obesity.

And last, can you guys give us a little bit of perspective about the SUMMIT study for patients with HFpEF?

Yes. So this was a study that we just initiated earlier this year, in April of this year. It is a study where we're -- it's a placebo-controlled study. We're looking at maximally tolerated doses of tirzepatide. So we have the maximally tolerated. We're trying to get the patients to the most tolerable doses, whether it be 10 or 15 milligrams, of tirzepatide. And then we're going to follow them for 52 weeks. And our end point in the study is actually a composite end point, both of a functional end point as well as a harder composite end point. So it's one where we have the composite end point of all-cause death, heart failure events, 6-minute walk test and then a KCCQ test from a PRO perspective. So it's a -- looking at it from a win ratio, we go through one to the next in terms of achieving the end points. And we're excited about this because it does provide the opportunity to see some of the hard events that occur with the -- with patients with heart failure and, at the same time, provide some of the subjective measures that you get through a PRO.

Okay. And the early thinking about why this should work comes from?

Yes. I think the focus there is again on -- in the weight loss that you lose and around the adipose tissue. It's an area that is impacted around the heart as well. And so an opportunity for patients who lose their weight will give them the opportunities to see these benefits.

And specifically in the heart fail population, I'm just not too familiar with the data that supports this. Is there like scientific data that you guys have that basically says, look, ex -- observationally in the patient loss ex weight, we are seeing that level of decline in MACE events. I mean is there anything -- what's the scientific underpinning for this logic?

Yes, I would point to the bariatric surgery data that's out there in terms of the amount of weight loss that is achieved there and the outcomes that come with those patients.

Yes. Got it. Very good. Jamie, Jake, we are at our end of our time, so I would really like to thank you for being here today. Highly educational as usual. And hope to see you and kind of wishing -- looking forward to seeing all your readouts over the next 12 months. Have a good day.

Sounds good.

Okay. Thank you very much.

Thanks for having us.