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

Okay. Thanks for joining us. I'm Tim Anderson, I'm the global large cap biopharma analyst here at Wolfe Research, and it's the second day of our health care conference. And we've been interviewing different members of management from different companies in a fireside chat format. We did 6 of these yesterday. We've got a couple today. The session right now is with Eli Lilly. And we've got 3 different folks. We've got Jake Van Naarden, who's the President of Lilly Oncology, who came to Lilly from the Loxo side when Lilly bought Loxo in 2019. We also have David Hyman, who's Chief Medical Officer of Lilly Oncology, also came from the Loxo side. And on the neuro side, we have Mark Mintun, VP of Lilly Neuroscience R&D. And Mark has been with the company since 2018. So as you can guess from that background, we're naturally going to talk about 2 disease areas, neuro and oncology. And we're going to start off the discussion today on the neuro side, partly because Mark has to hop off a little bit early.

So good morning, and thanks for joining us. Why don't we just jump right in? So we're going to, first, focus on Alzheimer's here, Mark, with you. So even though we have Aduhelm approved, still outstanding is the million dollar question, does plaque reduction lead to clinical benefit? We won't have those answers until we get Phase III results from different companies. Roche has that data in the second half next year. Eisai should have their data. And then I believe your Phase III data would come in early 2023. So until then, a little bit of a state of limbo, and the controversy is going to continue, but it's still possible that these Phase III trials can end up not showing a clinical benefit? And so maybe a question there is what would happen to donanemab? What would that mean for you guys if the gantenerumab or lecanemab Phase III data ends up being negative? It seems like it would be a real death note commercially. And it's not out of the realm of possibilities that, that could actually play out that way.

Yes. Sure, Tim. Obviously, we're excited looking forward to all of those Phase III trials. The key thing, of course, is going to be what exactly happened in the trials. Obviously, Alzheimer's trials have had an incredibly checkered history of having a little bit of unreliability. There's difficulty sometimes in getting a homogenous population. That's certainly the reason when we recruited for our Phase II pivotal trial, we recruited a more homogeneous population and used more aggressive use of biomarkers to do that. Our worry is that, of course, it is harder to demonstrate efficacy as you go to a more heterogeneous population. So one of the things that will happen is that we'll -- you want to look at if a trial does not meet efficacy, what exactly happened? Did, for instance, they get the amyloid removal that they expected? That's going to be one of the big ones. And did they actually enroll patients with the correct severity that they were looking for? Some of that is going to be judged on the level of biomarkers. How much amyloid the patients have? Are they -- some of them will probably have tau PETs or how much of the tau PET they'll have and then basically what their placebo decline is. One of the biggest difficulties in some of these trials has been when the placebo decline is unexpectedly slow, which means a recruited -- a patient group that's particularly early or it's unexpectedly late, very, very quickly declining, which would mean they actually recruited a group that was unexpectedly more advanced than they thought. So those things are tricky. And -- but we agree with you that a demonstration that clear amyloid removal in an early Alzheimer's patient would make things much more difficult. But we don't expect that. We see the field is moving. There's a consensus, and we're part of it in developing this data that there really is a link between amyloid plaque removal and slowing of the cognitive decline. So we're not expecting that, but -- and if we see it, we'll obviously want to dive into the data, and we hope that other companies rapidly will make that data available.

Can you remind us some of the timing of your Phase III results, those really should be in that early 2023 window?

Yes. First half of '23. As we've recently said, we've closed screening on that. And therefore, if you look 18 months later, we should be able to start taking the data cut and the data lock soon after that. So we're looking forward to having that data set. Now what's nice about our Phase III is that we are stratifying the patients by the amount of tau they have. So we are basically staging the disease as the patient goes into the trial. So we'll have both the high tau, as we call it, and the intermediate tau. But in our minds, with my oncology colleagues here, this means that we actually are able to say how much of the disease has spread and be able to categorize the patients in that and look at their efficacy separately.

Is -- and I know it's highly speculative at this point, but you guys are going to be filing for accelerated approval. I'm wondering if payers or prescribers are going to limit it. Let's say you get approval, if they're going to limit the use of your drug to your study population and specifically incorporating those tau thresholds, I would guess they probably would not because I don't think it's that precise of a measurement and it's that easy to do. But it could, in principle, narrow the population for donanemab relative to the competitors who never incorporated those tau thresholds.

Yes. I mean it's a good question. And obviously, we're looking forward to having that conversation first with the FDA as far as the label and then with payers. It's, of course -- I think what you're referring to is that it's been long pointed out that there are specific ways you recruit for patients in a trial that don't necessarily mean that the drug is not -- is only efficacious in that exact group. I mean we have -- in this particular trial, there may be limitations of MMSE, which is one way to stage patient. That doesn't mean that MMSE has to be done in every patient that gets treated. There are limitations on age, which is more for practical and in this case limitations on the way we enroll patients with tau spread with the concept that we were able to get a more homogeneous and more powerful trial out of that. So we think that that's a possibility, but we also point out that during this period of time, between accelerated approval and what we would call traditional approval, I think the field, as you pointed out, is feeling a little bit in limbo in the sense of waiting for those larger data sets. And I think we've already seen that having strong data sets, it makes a faster uptake. But then again, the other thing is that we need a better infrastructure for things like diagnosis and treatment. So I think all of those mean that once we're approved, we will -- assuming we're approved with accelerated approval, we'll be working on those aspects of getting the infrastructure up and running. And I think the full commercial value of this would be realized after traditional approval.

The differentiation with your program, there's potentially 2 different elements to it. It's how you've constructed the development program where you only dose for essentially a finite window to get a certain level of plaque reduction, then you stop. And to me, that's inherently attractive to every stakeholder, patients, prescribers, payers. And then also, you seemed to -- and this was obviously side-by-side comparison, fast and deep plaque reduction. And so if that's the premise for how these drugs are working, then your drug looks the best so far based on the data we have, again, side-by-side comparison. When you're talking with KOLs, key opinion leaders, or physicians, or you're at meetings like CTAD, is that differentiation already starting to become appreciated, do you think?

Yes. I think it is. I think it's actually an interesting point because people are realizing that there was another way to look at this. I think there's sort of like the -- the simplest way is like you put somebody on a drug and you just keep him there. We don't see any target of our drug other than the plaque. And so when we demonstrated efficacy in our Phase II trial, our belief is that, that is strictly due to plaque removal. If there are no more plaques to attack, there's no point continuing the drug. That message, while originally people perked up and thought that, that was actually quite different, I think, were successfully sort of turning that dialogue in its head. It's sort of like, well, then perhaps the other way of looking at it is why do you continue to giving drugs if the plaque is gone? So I think the people that we're talking to, particularly at the CTAD, the conference that just concluded recently, do indeed begin to see this. Now they're asking reasonable questions, which is, for instance, does the amyloid stay down? And because we had people going off in our trial even after just 6 months of treatment because they had reached essentially 0 Centiloids of amyloid in their brain, we actually were able to monitor them, and we see the amyloid staying down. We also see things like biomarkers, such as our P-tau blood assay that goes down at 6 months and then stays down even after they go off drug. So we're seeing the continued benefit in both the tau pathology and the amyloid pathology after taking people on. And I think that supports our decision to have approached the dosing that way.

So your competitors would argue while you give it only for a finite period because your drug only hits plaque, it doesn't hit the soluble forms of Abeta. And they'll argue that you have multiple forms of Abeta that are relevant to clinical improvement, not just a positive plaque but still circulating Abeta. So is that the -- when we've looked at this in the past, the science supporting the toxicity of Abeta oligomers is quite thin. But how -- what's going to be the -- do you think the competitive response there? And the physician -- physicians thinking, "Oh, no, you do probably need to get these drugs chronically versus short term."

Yes. I mean it's an interesting question. But we have a somewhat straightforward look at it. We have a drug that only hits the plaques. If there was value in hitting something other than the plaques, in my thinking, we wouldn't -- there would be a differential efficacy. But it turns out that ours, which only hits the plaques that we can tell, removes those plaques rapidly, numerically has one of the best efficacies out there. We don't see any evidence that the additional suppose of hitting the oligomers is giving differential efficacy. Given that, we sort of agree with your concept that the evidence is a bit thin for justifying continuing somebody on a drug. So it's a tricky situation. Long-term data is probably useful. As we pointed out, we're already seeing up to a year after stopping treatment that the biomarkers are responding in a way that shows continued benefit. So we're not -- I think that will be a continued discussion in the field. But at the moment, I don't see it going anywhere else than it is right now, which is a thought, but not necessarily any data there.

Well, let's talk about the head-to-head study you guys announced recently, which is comparing donanemab to Biogen's Aduhelm. So head-to-head study, a smaller trial, the primary endpoint is looking at plaque reduction. I think the intent of the trial is obvious. But why don't you tell us what the intent of this trial is? What are you trying to show and why you run this run?

Yes. The FDA established that amyloid reduction, they're treating it as a surrogate marker for a reasonable prediction of clinical efficacy. We believe the data over the next 1.5 years will continue to support that. That's our assumption. That's one of the reasons we're going forward. And so having the ability to do a head-to-head on essentially that measure of efficacy, as established by the FDA, made sense to us because each trial is a little different. I just got done pointing out a little earlier that patients are enrolled, they have different levels of amyloid, different levels of tau, different types of levels of disease. That could affect theoretically the amount of amyloid removal. And so to truly be able to talk about our drug having rapid and more complete and deep amyloid removal, we thought that we want in that dialogue real data doing a head-to-head. So that's our primary point. And clearly, we will be getting other information on that. In particular, we will also be able to report information on safety and other things. But we think that the primary information on efficacy -- this efficacy measure of amyloid reduction, particularly at 6 months, is going to be, I think, an important addition to the story that we'll be telling.

Head-to-head trials are not done that frequently, unless it's a regulatory requirement regardless of the disease. And the reason they're not done that frequently is you're essentially rolling the dice. If you hit it, great. And if you don't, you may have to repeal your product. So here, you could argue that donanemab has differentiation that stands on its own without the need to do a head-to-head. And I'm sure you guys have thought about, is this an unnecessary risk? So you could pause it, for example, that with really fast and deep plaque reduction, maybe at higher rates of ARIA, you destabilize the vascular membrane and you can have higher rates of ARIA. So what if that ends up being a finding that comes out of the head-to-head trial, wouldn't that potentially be a problem? And you must have confidence around not having that issue, my guess, otherwise you wouldn't be doing this trial.

Yes, I think you've summed it up appropriately, the types of thinking that you have to do to do that. We have a little bit of an advantage because we have so many different biomarkers on our patients. We can characterize them quite completely and stratify by the levels of pathology. So we believe we can -- we have probably a better-than-average ability to predict the risk of these type of trials. But the other side of the coin is understanding and having more data on ARIA is, I think, right now, an important concept for the field. And I think all -- we hope not that has to lead the way, but we will be paying more attention to thinking about ARIA and being able to get data on that. This would just be one part of that.

I want to shift to a discussion about subcutaneous dosing. So now the program that stands out as differentiated is Roche's because, at the moment, they're the only company that would launch with a subcu product. And you could also see how that would resonate with patients and probably lower health care costs overall if you're able to stay at home and get your drug that way versus going into an infusion clinic. So a question for you is the viability of subcu. In my past discussions with you guys, with Dan Skovronsky, for example, I asked the question, why not pursue subcu with donanemab? And he articulated that originally, I think, you guys were not convinced that subcu would get you high enough blood levels of drug to effectively get across the blood-brain barrier and lead to plaque reduction. So can you just kind of talk about this potential issue? And I wonder, frankly, that's a liability with the Roche program as maybe they don't get the level of same plaque reduction that other compounds get or maybe they get it slower. So essentially, they hit it, but it takes a lot longer to get there. So how is Lilly viewing subcu dosing today based on all the data that you've seen? And as part of that discussion, you can talk about your follow-on compound where I think you are pursuing subcu formulation.

Yes. Yes. We don't see it as -- as the data accumulates, we don't see subcu dosing as being incompatible with high amyloid removal. And we obviously want to pursue that as a possibility of this compound we call N3pG4. It's our -- as you pointed out, it's our sort of follow-up compound. We think we're seeing indications in our Phase I that we can achieve the type of levels necessary in the blood. But the -- at this point, that remains our approach to the subcu formulation concept. We still don't think donanemab is very practical given where it is in its development to pursue with donanemab. But if the -- our follow-up compound and things go well, we're hoping that we can be -- being able to progress that rapidly into Phase III with a subcu formulation. So we think that it is -- we've done our calculations. We've looked at this, and we think that's an actual possibility, but we still are going to need to have more data before we start that.

And you and I have talked in the past about what potential regulatory requirements would be if the company were to try to convert an IV formulation into a subcu. So we know that Eisai is attempting to do that. Biogen is doing that, early in their efforts. I thought and I think maybe you'd agree, but give us your current thinking. You probably can't run simple PK/PD trials. You probably have to run an actual Phase III trial maybe with the endpoint being plaque reduction if you were a company that had an IV formulation to start with and wanting to convert that to subcu?

Yes. I can't -- and I can't remember what I said exactly before, but I haven't had a conversation with the FDA on this because that's not our current pathway. So I can't be speaking from a regulatory point of view. But it would be our assumption that given the incredible importance that the FDA has put on to that specific surrogate endpoint of reducing amyloid and how straightforward it is to get once you do those PK studies, I'm assuming that, that would be a logical thing to request or require. It seems straightforward that the PK would not be good enough to directly predict the amyloid removal rates. So that would be my thinking.

Last question, then we'll move on to oncology. In NCD, for instance, was watching those closely trying to figure out how it will play out related to Aduhelm, which has pertinence to the other compounds in the class. Lilly's view of how the NCD will play out? And specifically, what sort of restrictions are likely? It seems like you could potentially have coverage. In fact, I think in my opinion, CMS will decide to cover. I'm trying to figure out what the restrictions will be. Do they knock out patients based on all the inclusion and exclusion criteria, trying to risk minimize -- trying to minimize the risk of ARIA, so maybe you take out hypertensive patients? I think the exclusion criteria, you didn't allow patients that had angina or MI. So how are you guys thinking about the NCD coming up?

Obviously, we realized that the full range of options are still on the table here. We look forward to January. We will start getting a look at the draft NCD. I think I agree with you that a lack of any coverage is probably unlikely. I think there is an option that I still think would be very, very difficult, is coverage with evidence development. That's still coverage, but it does dramatically restrict the way that patients can get treatment. And it restricts it perhaps in a way that makes it particularly complicated and actually makes access for patients and finding providers that have signed up for the CED difficult. So we're hoping that they find a path forward with true approval. The data, of course, that they have in their hands is not a lot. We've published our efficacy data in a peer reviewed journal. And -- but that's not true for the one that's already on the market. So I think they are a little bit between a rock and a hard place, as you might look at it. But my sense is that they will come up with coverage. I think they'll respect the need that this is an incredibly important drug and needs to be an option for patients. The exact way that they limit it to try to protect patients from a safety would seem to be difficult if you start hard coating that in an NCD, which takes a lot of work to revise when there's not a lot of data that exists on exactly how we can stratify that risk. I think the field needs to collect more data on that. But given our current situation, I would be a little surprised if they thought that they could put a very prescriptive coverage in there rather than let physicians choose for themselves. I think we will be ready to work with them on CED, and hopefully, they will not restrict it from the point of which -- too highly which physicians prescribe it as far as the evaluation. But I would not be surprised to see some level of restriction on the types of physicians that can prescribe it due to the need of evaluating the patient's clinical stage as well as being able to follow the patients under treatment.

Great. Okay. Well, thank you, Mark, for that discussion. We'll shift over to oncology. Jake and David, thank you for your patience. Let's -- we'll start by talking about Verzenio in the CDK4/6 class. The key to me is, is this really a differentiated drug? There's 3 drugs in the class in major markets. And you could argue -- I think the data very much suggests that maybe there is differentiation here, and that your product is different than Ibrance and -- versus Kisqali. We saw that early on, I think, even with the safety and tolerability profile being different. And then, of course, you could argue it's there because you guys hit adjuvant with Verzenio and Pfizer did not get adjuvant. So those data points raised the specter that maybe Verzenio is a different drug within the class. To me, a key point that will help prove that will be this upcoming set of results from Novartis on the NATALEE trial. So we know that Pfizer failed adjuvant. We know that you had adjuvant and Novartis is going for adjuvant. And Novartis' drug to me looks more like Pfizer's drug than it does Lilly's drug. So I know all these adjuvant trials are all constructed a little bit differently. But to me, that's really pivotal to kind of the future of Verzenio. Because if Novartis fails with NATALEE, it would then suggest that maybe your drug actually is mechanistically differentiated. So how are you thinking about differentiation and the importance of NATALEE to kind of proving that argument?

So I think the data that we've seen so far, the actual data we've seen across all the different treatment settings, to our view, show Verzenio abemaciclib to be a differentiated product relative to the other 2. And I'm sure there's differing opinions from different physicians out there on that topic. But to our view, it's a differentiated agent by virtue of its clinical data. And there are some preclinical attributes we could point to that potentially lead to that, though that's a bit speculative. We don't spend a ton of time worrying about the readout of NATALEE as it relates to sort of the future state. I think if NATALEE fails, certainly that maybe plays into the argument that you're espousing. Even if NATALEE is successful, though, it doesn't really change our trajectory in a meaningful way, in my opinion. It's a very different treatment regimen. I think most physicians are not that interested in 3 years of therapy with one of these agents. And the trial design is meaningfully different, forget even about 3 years, than monarchE was. So I don't -- we have to execute on our plan, educating on the labeled indication that we currently have in the adjuvant setting, educating on the label indications in the metastatic setting, continuing to grow share there. That's where our focus is, and we've got a great opportunity ahead of us to really help patients. We don't spend a ton of time teeth gnashing over the outcome of NATALEE. I don't think it has a meaningful impact on our trajectory, honestly.

Yes. And Kisqali stands out as having a unique toxicity on the QTc front. So remind me, even if they hit with NATALEE, you still have a drug that's kind of wounded relative to the 2 other competitors.

And the 3-year thing, Tim, as -- I mean you could talk -- they're not interested in it.

Yes. Okay. Pfizer gives assurances about the stickiness of Ibrance in metastatic. So Ibrance, biggest drug in the class, by far the market leader, probably because they got there first, well ahead of anyone else. They're the only company though to formally not have an OS benefit, formally in Phase III. So they've done these real-world analyses that show they have it. And that's always raised the question, will they gradually leak market share in metastatic to those that actually have shown a formal survival benefit, namely you and Novartis' drug -- your drug and Novartis' drug? So far, it has been a sticky market, and we haven't really seen Pfizer lose market share. So can you just kind of talk about commercially what you're seeing? And is it going to be a sticky market where Ibrance likely remains the market leader in the metastatic setting?

Yes. I don't -- I would characterize it as neither sticky nor not sticky. In other words -- I mean, first off, kudos to them. They got to market first. They ran an excellent development program, and they've commercialized unbelievably well. They've established this class. We wouldn't be where we are, frankly, without them having paved that trail. And as a result, they've borne the fruits of that commercially in continuing to be the incumbent and having dominant share. Now that having been said, so we were third to market, years behind them. And like just right now, we're at about 30% share of new patient starts in the class. So that's pretty -- I mean it's not 70% share, but 30% is not nothing. And we were around 26% to start calendar '21. To start calendar 2020, we were about 20%, again, all on the same metric. So we've steadily grown it. I think growing it from 30% or low 30s beyond that in the metastatic setting from here, I think I'd like to see us do that. It's going to be hard. We don't have a lot more data to talk about in the metastatic setting. So it's always hard to grow your share when you don't have more data cards to flip. So that having been said, we can continue educating on the data that we do have, and we will do that. I think there is headroom above the 30% where we are now, but I don't think we're going to double the 30% anytime soon. And that I think does speak to the stickiness of Ibrance prescribing. But I'm proud of the fact that we've grown share the way we have. I think it's a testament to our team, but it's also a testament to the clinical data we've generated on the drug. Obviously, the next leg in the story for us, obviously, is launching flawlessly in the adjuvant setting, hopefully over time expanding the indication to the entire ITT population of monarchE and then, of course, the prostate development program for the drug.

Is there -- so just finishing out the discussion on metastatic share. You could argue that adjuvant uptake will have a positive halo effect in metastatic because more and more physicians will start using your drug in adjuvant setting against the only available product at this point. So that kind of, to me, really has to help because you'll just get more users of your drug. And if you're using in an adjuvant, they might say, well, why not use it in metastatic as well?

I think it's a fair point. You're invoking something we call like the nonusers. And there is a decent subsegment of physicians who self-identify as nonusers or we could identify them based on prescribing data. You're right, given we're the only option in the adjuvant setting, the nonusers, so to speak, will start to get, we expect, experience with the drug, experience managing the diarrhea. The degree to which that then translates to a comfort level and a willingness to prescribe in the metastatic setting is just unproven. I hope you're right. We all hope you're right about that thesis. I can't lean into it right now because I just don't -- haven't seen it yet. So -- but it's -- it would be nice tailwinds, obviously, for the overall brand.

Moving to adjuvant with Verzenio. So you've got a label now. It's limited to those that have the high Ki67. And you sized that market out, I think you mentioned this in the past, 8,000 to 10,000 patients. Is that a U.S. number only?

Yes, that's a U.S. number.

Okay. So you're continuing to look at monarchE in the entire trial population and you need to show a survival benefit to get approval there, it sounds like. The size of that...

Just -- I mean survival benefit, I don't know if I would go that far. That sounds like statistical significance, which is not what we're talking about.

Okay. Got you. So you may not need to formally hit OS?

We can be more specific. We published data in the Annals of Oncology as a supplement actually to our original publication about a week later. This is now 3 weeks ago or so, you can find it online. We showed the survival curves for the overall population as well as the Ki67 high population that were indicated for. So you can see the survival trend that the agency was comfortable with in what's now the indicated population. And again, I'm not saying that's their bar. I don't know what their bar is, but I can just tell you that's what they -- that was the basis of their approval decision. You can see that in the public domain.

Yes. Okay. The size of the entire population -- the label population is 8,000 to 10,000. What's the broader population of monarchE?

So the 8,000 to 10,000 is about -- we calculate it's about 40% of the total.

Got you. Okay. And then there's been some questioning about the rate of uptake in the setting just because you don't frequently measure Ki67. So how has it been impacting...

That's not quite right. So Ki67 is frequently measured. It's not so much getting the testing to happen, it's the linkage between the test outcome and a treatment decision, which is a new phenomenon for the pathologist-oncologist interaction on this particular biomarker. It is frequently tested for, but it's oftentimes tested for in the context of both like diagnostic confirmation and then just prognostic importance so that a physician can sort of gauge a patient's expectations about their risk status. It's a slightly different framework to the thinking, "Oh, this is now a treatment change I have to make." And that's the education process that we're engaging in. It's not trivial. I mean this is -- this is -- this community hasn't had a change in -- like a meaningful non-endocrine therapy change in treatment in like 20 years in this setting.

Okay. I want to move to your pricing decision and announced strategy on Tyvyt, which is your PD-1 product, and then we'll shift to some pipeline questions. Hurdles in the U.S. to adoption, they're fairly clear. You guys have been fairly transparent. This strategy will work for certain segments within the U.S., maybe closed systems like a Kaiser, for example. What I haven't heard you guys talk about in your formal announcement, it was outside the U.S. So to me, that may be where the better opportunity is, in places like Europe where you don't have buy and bill and you don't have the same payer dynamics. And so there, where governments in certain markets more often actually influence what the physician uses, which doesn't really exist as much in the U.S., you could actually make more hay with a lower price strategy in Europe. So my question really is, is that an accurate assessment?

Yes, a couple of things I'll say on that. So number one, we don't know if the drug is approvable in Europe on the basis of the data that we have. We actually don't yet know if the drug is approvable in the United States based on the data that we have. That -- those are 2 important caveats. I think your point about the mechanics of European payer bodies is right. However, the idea here is to be disruptive. There's more headroom to do that in the United States based on United States' pricing for the branded agents as it stands today. And so there's just -- there's less delta in Europe essentially to offer to the system. And we're offering a drug without nearly the breadth of indications that the existing agents have. So you're right on the mechanics, but when you -- I think the details matter here, and they're not -- it's a little trickier.

Okay. So you just said you don't know if the drug is approvable. And that -- wondering if that speaks to the other question I want to ask you. I think the question everyone has, not necessarily limited to your product, but are China-only datasets adequate for approval in the U.S.?

We're about to find out.

So that's what you're alluding to, it's that question?

Yes.

And what's your thoughts on that. U.S...

My thoughts have changed over time. When we made the decision a year and change ago to get access to the ex China rights for this product based on the sort of political winds at the time, we thought that there was an appetite for this. In sort of quasi public settings at various conferences and the like over the past 6 to 9 months, we've heard statements that have suggested otherwise, that maybe the political winds have shifted. And obviously, there's larger geopolitical forces at work between U.S.-China relations that could be invoked here that are certainly above our pay grades. So I don't know, is the short answer. We don't know. We have an ODAC coming up. We don't actually know the topics of what the ODAC will discuss, but I don't think it's overly speculative to suggest that this will be perhaps one topic, but we don't know that.

And then -- so your answer is really a U.S.-focused answer, I think, because you're talking about U.S.-China relationship?

That's just the most advanced conversation we've had with regulators.

Yes. But in Europe, you don't know if that same...

We don't. I don't know.

Okay. Okay. Well, it will be interesting to watch how that plays out. All right. Let's move to the pipeline. So pirtobrutinib, LOXO-305, fantastic efficacy. And looks like if it was on the market today, it would probably be a market-leading BTK inhibitor. But there's naturally some steps you have to go through to get to market, in the big money indication for front line. And so you've kind of come up with a strategy to try to get the product to market more quickly, and it's not solely dependent on the head-to-head versus IMBRUVICA. But now you're looking at this BRUIN CLL-313 trial going after Rituxan and Avastin. So can you just kind of talk about what that entails? How much time does that shave off? It still looks -- if I believe clinicaltrials.gov, it still says -- shows a primary completion date of 2024. That's probably going to be quicker than head-to-head versus IMBRUVICA, but it's still years into the future. So if you can just talk about that? And then also, are there other things you could do that could actually pull forward an approval time line in front line, for example, change the endpoint, get approval on something like MRD, something like that?

Dave, do you want to take this one?

Yes, sure. Well, I guess, let me just start off by saying we agree with you that we're really excited about the efficacy and profile and proposition for pirtobrutinib today. You kind of -- the framing of this question kind of implies that the real long-term opportunity for this drug is as a first-line BTK agent. And we certainly want to enable a development program that allows physicians to make that decision, both from a prescribing perspective but also enables them with data that justifies that decision. In the near term, though, we think there's a huge opportunity for patients who are relapsing after treatment with modern agents like covalent BTK inhibitors or BCL2 inhibitors. We haven't spoken specifically about the opportunities for approval based on our Phase I/II BRUIN study. We're going to give an update about that at the R&D Day later this year. That's obviously a complex -- an ongoing discussion. But we've also enabled a series of randomized experiments in relapsed disease, which also have -- one of which also has near-term approval opportunities in it. I think in terms of first-line prescribing or really just line agnostic prescribing to allow physicians to use the drug wherever they feel it offers the most benefit to patients, you've kind of keyed in on 2 different programs. The first program that we decided to launch initially was the study against bendamustine, rituximab, which still gets about 25% market share as first-line use in the United States for newly diagnosed CLL patients. And as you alluded to, that's just by virtue of the outcomes in the control arm, a study that reads out more quickly. We also think to make a meaningful impact in first-line prescribing. Physicians are going to need more data than that. And that's why we have spoken publicly about our intention to launch a head-to-head study against ibrutinib, IMBRUVICA. That's, we think, a really important study for our molecule and an important study for the field, but it's a long study. So that's a study that doesn't read out in a couple of years. And I think when you want to really change and advance the field in CLL, you're talking about studies that take years because of the outcomes of that disease. Just the last thing I'll just say, as we talk about kind of not only the patient opportunity, we talk about kind of the market itself from a commercial perspective, we see overall a trend away from just monotherapy continuous prescribing as the kind of universal treatment of choice for newly diagnosed patients to time-limited treatment options as well. And kind of as you look at that landscape with more patients in the first-line setting, electing for time limited options, it actually moves a lot of the prescribing time for BTK inhibitors, including pirtobrutinib into the second line and beyond. So the idea that the "first-line market is kind of the biggest market" and where the majority of the value is for patients and for companies, I think, is likely to change over time.

I want to be sensitive to your time here because I know you have another meeting coming up, but can we just touch on your oral SERD. You've moved it into Phase III. I have 2 questions. The Phase I data showed at ASCO, good activity at 400 milligrams, no activity beyond 400. Unusual, I know. It's a small number of patients. But it really raises the specter of do you really kind of know the right dose to take into Phase III. So if you can just talk about that dose response data. And the second question on your program, is this trial as the primary -- is the statistical analysis plan really just going to look at ESR1 here?

Dave, why don't you take both of those, and I can round up.

Yes, sure. I wouldn't overinterpret the overall response rate across different doses. Obviously, we continue to treat patients and learn more about the drug. I think, in general, the broader idea here is that single-arm data sets, in particular, RECIST response rate in ER-positive breast cancer are very challenging to interpret. We obviously have looked at a variety of endpoints, both time to event-based endpoints, ORR in individual patients. We've done -- we have a window of opportunity study looking at the 400- and 800-milligram dose. So we're comfortable with our dose selection, both on the basis of target coverage, efficacy and safety. So kind of more to come, and I think that will be borne out in future data presentations. Your second question was about -- remind me, sorry, your second question was about...

Looking at ESR1 and the...

Yes. No, our -- so it's an important question. Our study is designed to look at the effect size in the overall population. And unlike the EMERALD study, which actually established the percentage of patients with ESR1 mutants and ESR1 wild type status, so basically it didn't enroll a prevalent population, but it declared the number of patients in each group, we're enrolling a prevalent population. Obviously, we'll look at efficacy within the subgroups. We think for an oral SERD to be a meaningful product, it needs to offer benefit to a wider swathe of patients than just the ESR1 mutant subset?

Jake, did you want to add anything on that?

Yes. I just -- yes, I would just really emphasize Dave's last point. I mean if the oral SERD class is really relegated to an ESR1 population, I think that limits its broad utility significantly. And particularly if the ultimate natural fit for these products for patients is in the adjuvant setting, a drug that really only works relative to existing endocrine therapy in ESR1 mutants is not that relevant for the adjuvant setting.

Do you think that's low risk? I mean do you think that will -- do you think these drugs will work more broadly as a class beyond ESR1?

I don't think we know. I think we'll get our first hint of that, looking at the EMERALD data in a couple of weeks. But the flip side is I don't think it's at all surprising to see a meaningful effect size in ESR1 mutants. I just don't think that's really the question.

Yes. Okay. Great. I want to let you guys out. Thank you very much for your time. Thanks to Mark as well. I think he's gone. Thanks to the IR team. We look forward to some of these data sets coming up for you guys. Have a good day.

Thanks, Tim.

Thank you.