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

[Audio Gap] for being here. Pleasure to have management team from Eli Lilly. There's a ton to go through. But Dan, I'll turn it over to you on what's on top of your mind, and we'll jump right into it.

Yes. No. Thanks for hosting us here. It's been a great meeting so far, and nice to have this opportunity to talk at a fireside. I guess probably most people are excited about Lilly because of what we're doing in obesity and Alzheimer's disease. But just a reminder that we have a great business in immunology with a couple of new products being planned to launch. We're also doing great things in oncology with Verzenio as a huge product and Jaypirca, which is in the early stages of launching across indications. So really a very diversified pharmaceutical company even though I know we're going to spend a lot of time on the first 2 topics of obesity and Alzheimer's.

Sure. Okay. So I know there's a ton to go through, and I'll intentionally keep the questions much more R&D specific rather than more commercial in nature even though I had forgotten, but I recall last night that you do have some commercial responsibilities as well now. But maybe just so we're all on the same page, there's constantly this question that some things coming online in North Carolina, but we don't really know what exactly that is. And is it like modular or how that works? Maybe it speaks to a lack of understanding on how some of the manufacturing build-outs actually work. But -- and I realize that's not what you work on day to day.

That's fine, yes.

But what's the high level of -- is it like 500,000 pens being made in Indy and then North Carolina will do the same at some point?

Yes. Thanks for that question. And of course, I think everyone is keenly interested in the capacity situation for our incretin business, particularly here, you're highlighting the device assembly part of it, which is one component in the supply chain. Of course, there's API, there's sterile fill, there's device manufacturing. We haven't commented on like unit volumes per site. And so what we did say though was sort of a doubling of capacity potential with the North Carolina site coming on. Of course, capacity is not exactly the same as supply, and we've got to work into that to produce it. And that's just one stage of the process that we're talking about. The North Carolina site is a large site that is coming up in stages, as you suggested. And over the course of the next couple of years actually, it will continue to improve in its capacity potential.

Over the next couple of years.

Yes. And there's a sister site that's coming up also in North Carolina with a little bit of a longer time horizon that will add another increment of capacity. Of course, we're excited that the demand for incretins, injectable incretins is so high, and we'll continue to invest robustly to improve the capacity to manufacture these.

Got it. And then I don't -- like I said, I don't understand how the staging of these manufacturing capacities actually plays out in practice. But is it reasonable to assume 1/3 of whatever the total is could be online in first half '24, for example? Or it's hard to even quantify that?

Yes. I think it's hard to quantify. There's 2 ways that capacity improves over time from a given facility utilization. So one is adding additional lines of operation over time, which can happen by fitting more in the existing footprint, also expanding facilities. The other is within a line, we can optimize performance and get more unit volume out of it as we get more proficient at running it or things like that. So it's probably not accurate, just sort of saying what fraction of capacity comes up at a certain time.

But as you understand it, it was the pens that was the bottleneck from a Lilly perspective, not the API.

That's correct. That's the current bottleneck. Of course, I'm quite confident that we can find bottlenecks anywhere in the supply chain for incretins and whichever one we sell first, and the next one becomes equally important itself. So we're working to increase capacity against all 3 aspects of manufacturing, which is API, the sterile fill and the device manufacturing.

Got it. Okay. So maybe let's go into maybe the first topic from a more R&D specific. I know there's a lot of focus on oral GLP, specifically as it relates to safety profile. Could you sort of characterize for us how you're thinking about liver safety and also maybe even higher level, how are you thinking about overall risk on this program relative to some of the other GLPs that have moved forward?

Okay. Sure. Interesting framing to put it relative to other GLPs. But let me just start with the overall risk profile and particularly focused on liver since [indiscernible] question. One of the common ways that small molecules die in drug development is drug-induced liver injury, which is often difficult to predict. Maybe it could go stronger. It's impossible to predict from preclinical studies. It's -- not correlate with what we see in animals, and it manifests in humans as sort of sudden and severe liver injury. We haven't seen any of that with oral orforglipron, so we're really glad. We know though that a competitor saw that with their small molecule GLP-1 agonist. And that's caused sort of increased scrutiny over the liver profile of the molecule. What did we see with our molecule in Phase II? We had 2 Phase II studies, one in obesity, one in type 2 diabetes published in New England Journal and JAMA, respectively. And we detailed the liver safety findings in those 2 publications. First, at a group level, what we see is patients on drug actually have a decrease in liver enzymes. So that shows...

Group level.

At a group level, the average ALT/AST declines in a dose-related manner. That's probably related to driving fat out of the liver and improved liver health. And then there are a couple of patients noted that had excursions of liver enzymes above the normal range. That's not unexpected, particularly in a type 2 diabetes population where there's a very high level of comorbidity of NASH. And in all of our type 2 diabetes trials across molecules, we can see those kinds of excursions in liver enzymes. Importantly, the patients on drug, there are patients on placebo and drug, I think, in this program who've experienced it, but the patients on drug returned to normal without cessation of therapy, which is usually not consistent with drug-induced liver injury. So when we saw the evidence, we felt pretty good with -- we actually felt quite comfortable, I should say, with the liver safety. Of course, when the competitor data came out showing a liver safety injury, we looked again. We came to the same conclusion. Our Phase III program is going well. We have a lot of patients enrolled and continue to follow these patients over time with no signals here to be alarmed about. Of course, we monitor liver safety in all of our Phase III programs, and we're doing so in orforglipron, no different than the other programs.

Then I guess one question that I can answer for myself is when ALTs go above 10x for a patient in a trial, I'm not a clinician, but I would have thought I would pause whatever medicines they're on. I noticed in this specific instance, on your oral GLP, they dose right through it. Like how does that play out practically?

Yes. Different trials have different stopping rules for liver enzymes. Sometimes there's differences in time of when the investigator sees the labs or things like that. Honestly, I can't recall what the stopping rules were in Phase II and whether they -- the protocol asked them to stop or not.

And these were one-off upticks and they came right back down? Or are they kind of sustained for a little bit and slowly came down? How did that actually play out [indiscernible] ?

I don't remember the time course. Like I said, it wasn't a particularly noteworthy liver profile. It just looked like background noise that we see in our trials a lot.

Got it. Was it reported to you when it happened in Phase II?

No.

On a blinded basis? It wasn't. So you only found out afterwards as and when the Phase II was completely done on a safety basis?

You're asking me personally or you're asking the company?

No, as you're saying, on a blinded basis wouldn't like our rate of ALT above 5x is like 5 cases now on a pooled basis across trial, wouldn't that information be getting reported?

You guys thought the Phase III.

Yes. Phase III or Phase II for that matter, but I guess let's...

Yes. So on the Phase III, we do -- our safety team, of course, has access to the pooled data on a blinded basis. If they saw something, that would be reason to show caution, but we're fine.

Got it.

Nothing like that is happening.

And I know there's been some confusion lately on the exact doses in Phase III. So Phase II, you hit mid-teens at 45 milligrams but also 35-milligram came about 1 point lower than that as well. My understanding is Phase III is not using 45 milligrams, it's about 3 doses below that. Can you just remind us of that?

Yes. Well, we've tried not to disclose the Phase III dosing. But I will acknowledge that in Phase II, we saw probably a saturation of efficacy at the 2 highest doses. There wasn't much room between them at all. And so I think that led to our determination of the right dose to take into Phase III.

Got it. Is it usual by the way? I was thinking back to how often do I see 3 different doses for an active arm in the Phase III. It's somewhat unusual, but I don't know if it necessarily changes the powering very much. They all work. How do you think about that?

Well, it's what we did for tirzepatide. So we're pretty excited about that profile because initially, when we designed the tirzepatide trial, there was a lot of uncertainty about tolerability and how much the tolerability would improve with titration. So that kind of led to having multiple doses. And then what we found when we launched tirzepatide is that patients enjoy the ability to actually manage their dose decisions with their doctor and find the right dose that gives them efficacy. That's actually a positive in [indiscernible] space. So I hope we can replicate that same kind of experience with orforglipron, which involves some manufacturing complexity because you're making more doses but it gives patients the options to titrate up. So different than other therapeutic areas where you might try multiple doses because you're not sure which one will work and you're going to pick one to commercialize. I think the intent here would be to commercialize all of the doses.

Got it. And maybe not to harp on these doses too much, but for the Phase III dose selection, one of the things that stands out in some of the early data, both preclinical and in Phase I, is that oral bioavailability is perhaps not so high on this. So I was thinking back to tiny numbers, but 27 and 45 milligrams in Phase I having similar AUC and 3 and 15 milligrams having similar Cmax, and then there's like a [indiscernible] paper talking about bioavailability in Europe. I guess I'm most asking that in the context of sure, there's still saturation from an efficacy response perspective, but could that simply be a function of bioavailability playing out there?

Yes, yes. Of course, we look at exposure related to response as well as dose related to response to pick our doses for Phase III. But you raised a good point. I think it depends on what you're comparing to. So if your yardstick is oral peptides, you would say it has excellent bioavailability orders of magnitude better. On the other hand, if you're talking about like typical small molecule orals, I agree, bioavailability's quite a bit lower. But compared to the alternatives here, which are either injectable peptides or oral peptides, up until now, I think this is a huge leap forward.

Got it. My last one on this. I know there's a trial you're running versus insulin, it's ACHIEVE-4 study. And there were 2 things that stood out about it. One, there's only 1 dose, not 3-dose levels. So I'm assuming it's the highest dose level. I could be wrong. The other one is this was the only trial that you've posted online, which has ALT and AST exclusion criteria, spelled out on clin trials. I wonder if that ALT above 5x exclusion criteria was implemented across the Phase III program.

I think that the liver enzyme inclusion and exclusion criteria are the same for orforglipron as they were for tirzepatide and as they are for retatrutide. I know the inconsistency in what we published on ClinicalTrials.gov is confusing. Sometimes it ends up on there. All the inclusion criteria sometimes they're not included. But that I wouldn't read anything into that. That trial is not different than the rest of them in terms of the inclusion/exclusion criteria.

Okay. Maybe the last one on oral GLP for me is there's so much focus on this program now maybe in part because of there's so much generalist interest in Lilly and they think of oral as the next thing after injectable. If that program did not exist tomorrow all of a sudden, for example, do you think it changes overall the GLP strategy that you guys are on and the commercial competitiveness?

Yes. I mean of course, we want all of our programs to succeed, but I don't think Lilly is dependent on any one Phase III pipeline asset. I think part of the investor interest could be what you said, related to generalist investors. But actually, I think the patient acceptability of once-weekly injectable is quite high actually, and we are seeing a lot of interest in primary care use. And on the other hand, I know there's a number of companies with oral incretin programs that are investable and biotech valuations. So that could actually drive a lot of investor interest. Maybe they're rooting for their success or something. I think we're going to be okay.

Got it. I did say that was the last one, but I did have one...

Okay. That's all right.

I thought -- I remember I was tracking when would the first oral GLP Phase III come out for Lilly. And clin trials used to imply January next year, and I think now it suggests April. I wouldn't think of oral GLP as something that would recruit slower, so I'm not sure what happened there.

Yes, I'm not sure either. There's actually a computer program that populates ClinicalTrials.gov from our protocols and from our scheduling software at Lilly and it looks at the dates but...

It must be ChatGPT.

Yes, maybe. I hope it's not AI-powered. But the trial actually is enrolling faster than expected. So I can't really explain why the date has changed a little bit on ClinTrials.gov.

Okay. All right.

They're within an appropriate range of accuracy before and now. So...

Excellent. Mike, did we miss...

Just one question on...

The fifth -- 500th question on oral GLP?

One question on toxicity. You guys conducted 9-month toxicology studies in nonhuman primates. The safety margin was very low, is below 1. Any reason why that was? Or...

It's some of the no observed effect.

Yes, the no adverse -- no observed adverse effect level was 3 mg per kg in monkeys, but the exposure multiples for the 45-milligram dose were like 0.77, 0.87.

I actually don't recall the details of that, but I know in general, what we find with high-dose incretins in nonhuman primates is weight loss. And so often, we have to stop the tox studies because of either weight loss or GI side effects. We know that weight loss is intended pharmacology, and GI side effects could be mitigated by dose titration. So often in the nonhuman primates, we can't get -- for any incretins, injectables or orals, we don't get multiples of NOAELs because of the expected pharmacology. But in general, the nonhuman primates don't have diabetes or obesity. So it's not an optimal for study in these drugs.

Okay. Excellent. Maybe let's talk about GGG. I don't know what the internal...

Well, that is the internal name. The external name is now retatrutide.

Okay. On GGG, the - some of the cardiovascular observations on adverse events, they could potentially be on target because of the glucagon receptor expression myocardium. How do you think about that and the overall safety profile on that program in particular?

Yes. I think most of the cardiovascular effects we see are probably GLP-1-mediated. But undoubtedly, there's some glucagon pharmacology, and we know glucagon alone can cause some increases in heart rate and improvements in cardiac contractibility. I think these are likely to be nonadverse to the drug, and they're within the range that are seen with other incretins. So for example, one measure of heart rate excursions is 20 beats per minute and how many patients got above 20 beats per minute increase in heart rates. That's a threshold that's commonly passed with incretins. So for example, I think that semaglutide label has nearly 1/4 of the patients having 20 beats per minute excursion. I think we have fewer than that with retatrutide. But there are some patients who see increases in heart rate that attenuates over time as it does with other incretins, including semaglutide and tirzepatide, where people have an initial increase in heart rate. And then as they stay on the drug, it comes back down. And we now know, I think, and the reason I mentioned semaglutide is we know for the SELECT data, that despite causing heart rate elevations in a fraction of patients, it results in an overall positive cardiovascular risk profile. So that's not a particular concern for retatrutide. But like I said for orforglipron, when a molecule goes into Phase III, this is the time that you look for safety signals in a large population. And that's one of the many things that we look at with retatrutide. We also look at liver safety with retatrutide just as we do for orforglipron, and we'll keep our eyes on all the safety issues.

What about arrhythmia? Because I thought that was -- or at least I recall now. Arrhythmia, there was a prominent safety table callout for that with the...

Yes. I think that the arrhythmia we saw were related to the increased heart rate.

Okay, okay. So they were tied to the heart rates. Okay.

Yes, I believe that's correct.

Okay. Excellent. Maybe the last one on this. I think there's early hints that there might be an interesting kidney clearance, EGFR increase potential here, not just stabilization but EGFR increase. What would that look like? Kind of 5-point EGFR improvement, something along those lines doable? Or is there a certain minimum EGFR you have to be? I'm just trying to think about the CKD opportunity with this program.

Yes, yes. I actually think this has a very significant effect on the kidney and maybe too soon to speculate on the numbers. But actually, when we -- when I look at the totality of the data we have with retatrutide compared to like tirzepatide, there's a couple of places where you can put your finger on something that's likely to be a glucagon-driven effect. I think the degree of weight loss, which looks likely that it will be superior to tirzepatide, is glucagon-driven. I think the profound decrease of liver fat to 90% is likely glucagon. And I think this kidney effect that is possibly a benefit for CKD is likely to be glucagon-driven. So those are all things that need to be explored in additional clinical trials for sure.

Okay. Any questions on any of the topics so far? I want to make sure everyone in the audience gets to ask as well. Okay. So maybe we'll keep going. And I know GLPs are the magic drugs right now, and they work everywhere. Is there an indication where you think they're higher risk than others? And conversely, what are the lower risk? And I'm talking beyond obesity, beyond diabetes now.

Yes. So I don't know if there's much magic here. But clearly, I know metabolism and obesity are bad things. And GLP-1s and, of course, particularly dual agonist like GLP-1/GIP, which is tirzepatide seemed to have a very important effect in improving metabolism, which could lead to benefits in a variety of diseases. So of course, diabetes and obesity is where we're approved right now. In the next year, we're going to have a readout in obstructive sleep apnea in patients with obesity. I think there's a lot of potential there. We don't have any data. We didn't do a Phase II trial. Pharmacologic approaches here are untested. But we know that weight loss with bariatric surgery or with lifestyle modification can lead to improvements in obstructive sleep apnea. So I'm excited about that. We have a heart failure study that will read out next year. I think again, having seen competitors' data, we're pretty excited about the profound effect that this mechanism could potentially have there. And then on NASH, we have a NASH study, Phase II NASH study reading out next year, which I think could also be important based on the amounts of fat we drive out of the liver. So those are all pretty important indications that are tied to metabolism and weight loss that I think we could get a benefit in.

I want to spend more time on heart failure, not right now but on my own and down the road. But one thing that did stand out to me is much of the heart failure finding so far are not on a composite outcomes endpoint with the evidence to date, and it's more on measures that approximate those things. How do you think about cardiologist interest in outcomes-based endpoints for heart failure rather than some of the data we've seen so far?

Yes. I think they could both be important. The sema data we saw had improvements in the quality of life questionnaire as well as 6-minute walk. But when you look at the outcomes, there were very few because it was a smaller, shorter trial. I think we have a larger -- a longer trial. Maybe we'll get more patients driven to outcome and be able to demonstrate...

And then if you want that, in your view?

I don't know. Of course, it's always better to have outcomes than not to have it. So maybe that's an easy yes. Of course, it's great. There's so little advances for this population though that I think we should be pretty excited about what these incretins can offer.

Got it. And on NASH, the question that comes up is liver fat reduction is clear, 70%. What about fibrosis improvement, your expectations there? And any prior evidence that points us to fibrosis benefit?

I don't know what the reversibility of fibrosis is with a metabolic agent. When I think about the pathophysiology of NASH and simple idea, I think it starts with ectopic deposition of fat in the liver, which then leads as it does in every organ to inflammation. And then that inflammation leads to tissue destruction, which leads to fibrosis. So I would be confident, I am confident that we can reduce fat, and that can reduce inflammation. Probably, we should be able to see a reduction in progression of fibrosis. But in terms of regression or reversal of fibrosis, that feels like a higher bar to reach. There's...

Right. But conversely, Dan, isn't it that if it's F4 patients, which you're not studying, then it's very hard to see them come back to F3. But F2 to F3 is a spectrum that biopsy to biopsy for the same patient could go back and forth. So as long as you're not getting much progression, you kind of end up with the fibrosis benefit the way it's defined in the trial while not necessarily getting...

Yes. You're raising a good point, Umer, which is that we kind of think of a quick glance, you'll say, "Oh, the biopsy is a gold standard, tells you what's really happening with the patient." But actually, it's a tiny piece of tissue that's read sometimes in ways that can vary from reader to reader. And we take another biopsy from the same patient, you get a different answer. And so if you've had lack of progression, that could look like reversion to the mean in a large enough study. I don't know if we have a large enough study to create that effect, but it's definitely possible. And for that very reason, even a reversal of fibrosis should be interpreted with caution in a trial like this. And I really think that we should -- well, I think Lilly has an important role, and we're making progress in advancing the field of NASH towards biomarkers. We don't love doing trials with liver biopsies. And I think the future will be noninvasive ways of monitoring liver function and NASH progression or regression, if possible.

Got it. As I was going through the SELECT data, which came out at AHA recently, we were trying to think about the North America, how that performed versus the overall trial. And we said, "Okay, let's go find CVOTs for other GLP studies." I know Mike made a nice slide around North America performance. And it looks like several other trials, I believe including the Trulicity trial, North America did a little less than what the overall trial suggested. I guess what was the -- do you recall some of the FDA interactions around how North America population gets evaluated? Or were they focused on the totality of the trial?

I don't recall that being a review issue for Trulicity. I think with respect to sema, obviously, that feels like a novo question. There could be a lot of reasons why different countries could perform differently. Probably the first and most important one is that there -- none of them are big enough to be powered for stat sig. And so the variability on a country-to-country basis is to be expected. Of course, the standard of care and what alternative treatments are available and used will always impact also the magnitude of benefit that you can have for a given drug. So that could be part of it as well.

Okay. Excellent. I'll spare the other question, which is more SELECT specific, which is not a Lilly question. We can talk about the early separation another time with a different company. But on the question -- on the trial you guys are running, SURPASS-CVOT, semaglutide is a comparator -- sorry, did you guys consider using semaglutide as a comparator?

Yes. So we're running this trial as -- which I think is a really bold idea, head to head against Trulicity against dulaglutide. When we started this trial, Trulicity was the leading GLP-1 in terms of, I think, both TRx and NBRx. And importantly, and this is still true today, Trulicity is the GLP-1 that demonstrated benefit into to mixed population of both primary and secondary cardiovascular risk prevention. So we're excited to have the opportunity to compare to Trulicity as a gold standard for efficacy in this population. I think the study is designed around noninferiority since Trulicity itself has a benefit against placebo, I think we do not infer to Trulicity that shows a benefit versus placebo. But I'm optimistic that maybe we could also hit superiority, which is clearly a secondary here. Having seen the benefits of weight loss, there's a pretty big difference in weight loss between Trulicity, which probably had very little or negligible weight loss in REWIND versus what we expect to see with tirzepatide which, I think, in trials of type 2 diabetes patients showed about 14% or 15% weight loss.

Okay. But the critical thing, which is what I was going to get at, in a sub-15,000 patient trial, the goal is noninferiority because it is the first head to head versus a GLP in that CVOT setting.

Yes. That's the primary endpoint. Correct.

Okay. Makes sense. Mike, anything else on GLPs or anything from the audience on GLPs?

Perhaps it's more of a commercial question, but how do you see maintenance therapy in the neighboring class? Right now, it's not really -- you talked about our study right now, but everyone assumes that patients aren't going to take -- once they've lost the weight, patients aren't going to take the drugs as frequently. Any thoughts there?

Yes. It's a good question, Mike. I definitely maybe caveat it by saying it's easier to predict scientific outcomes than it is to predict human behavior. So I'm not sure what people are going to do. We want to make sure they have options. One option that people have right now is they can change their dose with their doctor of their incretin. So both -- well, actually tirzepatide has multiple doses that's approved as we were discussing earlier, and I think patients enjoy having that control of their disease. In the future, I hope we can have other options for them, too. Maybe some patients will prefer maintenance on an oral, I don't know. Maybe other patients will prefer to start their treatment with an oral and if they don't get enough efficacy, switch to an injectable. We'll have to see. We haven't done any studies yet to test different maintenance regimes. But I look forward to exploring that area. I think it could be important for people to know what options they have once they reach their target weight. One thing we know that is not an option is coming off of the drug and what we see is for most patients, their weight gain returns if they come off the drug.

So there's 2 components to that, Dan. One is you come off the drug. Clearly, there's weight gain.

Yes.

But I feel like we have not seen those randomized studies track long enough to see whether you go all the way to baseline or do you go half the way to baseline. Do we have a sense for that?

I think you're right that we don't have enough long-term data of people who've withdrawn from therapy and becoming increasingly hard to do those studies because well, we know losing -- treating obesity is important, and take patients off of a drug that's working is a hard thing to do. So I don't know. My prediction is if you look at the data with enough granularity, you'll find subgroups of patients who are better able to maintain at least part of their weight loss and other patients who may revert back to baseline. I think the mean of the population, at least in the duration studies we've examined doesn't return to baseline. But surely, some patients must.

Got it. Have you guys -- as we sort of -- because there's multiple layers of that duration question. I'm sure there has been work down within Lilly just to understand Trulicity historic median duration, how that compares versus Mounjaro. And I do understand the Mounjaro question is a little more complex because it's loaded with confounders like there was some pair changes, supply shortages, so you may not actually have a true duration number. But has that been a question that's been studied at depth internally on what is the actual realistic duration? Because that can clearly drive what type of future demand to expect based on the...

Yes. I don't think we have great ways of studying it internally. Unfortunately, I agree with your comment though that looking back at the last year of Mounjaro sales is not the way to do it because we had patients on a $25 copay card and then that went away and then they stopped. And there have been supply shortages for us and I think also for our competitors. So it's a confusing scenario out there right now. But I do note that for type 2 diabetes, which is where we've had Mounjaro up to now, for type 2 diabetes, Trulicity had very good patient adherence and refill rates that were superior to the orals out there for type 2 diabetes. And of course, Trulicity was treating something that was mostly invisible to patients, which was their A1C. Here, we're treating something that patients can feel and see as well as giving them other kinds of health benefits. So they can feel the sense of hunger that often accompanies obesity. And with human on these drugs, that usually goes away, and people often report that the food noise has gone way, that eating is a choice now. They see this must be how other people feel normally. I think that's a powerful motivator for people to stay on drug. We've also heard -- this is anecdotal as well, that when they come off of drug, that returns. And even before the body weight comes back, they start to feel the hunger and the focus on food that they don't find favorable. And so that could be a motivation to stay on drug. So I think it's early, early days here in the class. My prediction is patient adherence will be high and resistance will be long because of the benefits that patients have with these medicines.

Got it. I know lately, there's been a big investor question around muscle issue. How big a deal is that realistically based on some of the feedback you hear from various sites that have been working with you guys, commercial feedback?

Yes. Actually, we haven't heard that it's an issue at all from sites or commercial feedback. Specifically, what you're referring to is the fact that patients lose some lean body mass along with their fat body mass. Probably they lose about...

So that's how I lost my muscle mass.

Yes. Me too. So probably, it's about a 3:1 ratio of fat to lean mass loss. That's actually relatively consistent, whether we're talking about weight loss from pharmacologically induced weight loss or bariatric surgery or intensive lifestyle modification, that when the fat comes off, some lean mass comes off as well. It's probably a combination of metabolism as well as utilization. So in other words, when people have more weight to carry around, their core muscles are stronger and exercise more. When they lose that, they're less needed. The question is, does that have any functional consequences? Actually, what we've seen so far is functional benefits to weight loss, again, whether it's intense lifestyle modification or bariatric surgery or pharmacologic, there's actually functional improvements because their overall body composition has improved. They've lost more fat than lean mass. So therefore, they generally feel like they can conduct activities better. However, there may be some patients, although we don't have evidence for who this is yet, that could benefit from both weight loss and maybe more muscle. And that was the thesis behind bimagrumab, and we're going to combine it with tirzepatide to see if we can drive weight loss while at least preserving, if not increasing, lean body mass. The challenge will be to show functional benefits from that, not just a difference on a DEXA scan that show that, that muscle mass really leads to even better function than the functional improvements we expect to see with tirzepatide alone. It's a long...

But wouldn't that be a very complex endpoint then? Healthy individuals knowing that Novartis myositis trial was more tricky on that?

Absolutely. I think regenerating functional muscle has proven not successful yet with this drug or with others. Maybe maintaining functional muscle is easier. But I know a lot of investors and questions have been around, well, is this correcting a liability of the incretin class? That would actually be easier to demonstrate a Phase III trial if it were the case, but that is not the case. It's more about trying to add an additional benefit, which is going to be hard to prove. But that's why we bought Versanis, we're ready to take that on.

Got it. Dan, so one of the -- my impression of Lilly's R&D strategy, your R&D strategy over the last few years has been there's been constantly this having iterations of various molecules on that validated target, oral higher efficacy, different second target, et cetera, which has been very successful. One thing I've not seen enough of or one thing I'm starting to see from other companies so far is novel non-incretin orals going to the clinic. So Amgen has one, and Pfizer has one as well now. I guess where is your head at on that? And is that something you're going down the direction of as well?

Sure. I would say rich in all mechanisms for obesity. I think it's probably wrong to assume that just because the incretins are working for obesity, that this has sort of ended the drought, and now everything that we try will work for obesity. I think the bar is really high for anti-obesity medications. And we'll be selective about what kind of mechanisms we test. Probably the most important factor to consider is safety. Incretin classes set a high bar for safety. And I think there can be very little tolerance of safety concerns for anti-obesity medications. The second consideration is efficacy, and the incretin class has set an extremely high bar for efficacy. So it's been traditionally hard to make new mechanisms for weight loss. I expect that hasn't changed. But I know investor interest has changed. That's great. We welcomed more investor this year.

I want to move past GLPs in the last 10 minutes, but there's 18 seconds remaining. I will ask you one follow-up. Is there more intensified liver monitoring on your oral GLP Phase III? I think there was some confusion around what you said on the call.

No, it's...

It's the standard...

To my knowledge, that's the standard, same as tirzepatide, same as retatrutide.

Okay. Got it. Okay. Excellent. So let's move on past that. I know for abemaciclib, there was a trial due in prostate setting, which could have been sort of well beyond breast cancer. Is that prostate trial, CYCLONE 2, still due late '23? It's event-driven. I don't know if there's good visibility on that.

Yes, it's -- as you said, it's a bit different trial, so we never know exactly when we'll finish. But I think late this year or early next year is the right target to think about for that trial. Pretty excited about that trial because we had a -- you can think of this as an adaptive design where there was a Phase II part of the trial, which we didn't see the data from, but I had a high bar to adapt it to a Phase III trial. So efficacy topped that bar, and then the trial expanded automatically to become a Phase III. So excited to know that at least in that first cohort of patients who also comprise part of the final efficacy dataset, they topped a high bar for efficacy that holds up with the rest of the population. There's some pretty great drug.

So the previously disclosed data for abema in prostate, I'm not talking about the Phase II that sort of passed the bar. The prior data was only 7% response rate, but that's not unexpected because of CDK4/6 mechanism, but also PFS was about 6 months, 6.5 months. I guess could you remind us, based on that alone, I wouldn't have said it as passed a high bar, but I know there was a randomized trial, which passed a certain bar. Can you remind us what the design was there?

Sure, sure. So second one, first of all, a small trial, maybe it was a Phase I or Phase IIa, which enrolled pretty late-line patients with prostate cancer. So really sick patients in tested monotherapy, which Verzenio works in breast cancer monotherapy as well, but it's pretty small effect size. I think the main mechanism of CDK4/6 inhibitors work is in combination with endocrine therapy. CYCLONE 2 is in combination with endocrine therapy in prostate cancer. CYCLONE 2 itself, this Phase III we're waiting for, started out as a Phase II trial. And then based on a positive endpoint in that Phase II portion of the trial, it adapted to become a Phase III.

And that Phase II was randomized?

Yes.

Okay. So that's critical. So CYCLONE 1 was monotherapy. CYCLONE 2 was endocrine add-on, and it was randomized and that's where you...

Yes. And I would imagine less fewer lines of prior therapy, less severe population.

Got it. Got it. Got it.

But probably the endocrine therapy is the biggest contributor to efficacy, I would...

Got it. Okay.

But we certainly wouldn't have started this Phase III based on CYCLONE 1 in the absence of CYCLONE 2 sort of interim analysis.

Got it. Maybe a program that's near to your heart, Alzheimer's program. How are you thinking about, a, sort of donanemab setup into the launch next year? I realize Biogen is progressing with their launch right now, but there's a lot of commercial challenges they're encountering. How are you thinking about sort of the market development? Is it not what you would have thought?

I don't know. Of course, this is near to my heart because we've been working on it for a long time. All of our programs are near to my heart. But in Alzheimer's disease, having worked on it for a long time, I've noticed that there's a pretty slow adoption of new ideas and new technologies across the field. I think neurology as a practice has probably not seen many advances in Alzheimer's disease. And so my expectation is it will take a little while for practice patterns to change around diagnosis and treatment. And this is a complicated class of medicines to use, in many cases, requiring a PET scan for diagnosis, which is a technology we developed and hope to see more uptake of now. And then a drug that's delivered by infusion in the case of our competitor every 2 weeks for donanemab, hopefully over 4 weeks if it gets FDA approval, which we're expecting early next year. So that's a new thing for neurologists that give a lot of infusions. And then there's a side effect called ARIA, which requires MRIs to monitor, and that adds more complexity. So it's going to take time. Most of the prescribers we talk to indicate a willingness to try these drugs, a belief in the efficacy, which is great news. I think that's huge progress that they think the effect size here is real and meaningful benefit for patients. And their main focus now is on safety and going slow as they start patients so they learn how to manage patients on these drugs. I think that's a healthy setup for long-term growth. But at the same time, I think that setup probably indicates a slower ramp than, for example, what we've seen with obesity drugs or something like that.

And Dan, how are you thinking about the ARIA rate findings on remternetug, especially on some of the doses that are competitive?

Yes. It's still early to say. I think this is kind of a dose-finding regimen part of the Phase III that will then go into a focus on a single dose as the trial expands. I think our goal here is to continue to do our best to drive ARIA rates down, particularly symptomatic ARIA. That's what we're most concerned about, particularly serious adverse event ARIA. The radiographic ARIA rates are probably a little bit harder to index on and depend on a number of factors.

Got it. In the last 4 minutes, I wanted to touch up on your new job, immunology. And maybe a couple of questions there. One, how do you see the IL-23 class playing out in the IBD space? I always felt like the early data on UC was very competitive versus ENTYVIO. We know ENTYVIO is very sizable. You guys have obviously gotten a leadership position timing-wise on the IL-23. So how are you thinking about the broader IBD opportunity?

Yes. So maybe I'll just start at a high level, which is IBD is a pretty significant disease. It's probably about 1 million people with Crohn's in the U.S. and less than that with UC, but it's still quite large. Biologic penetration still isn't where it needs to be, particularly in UC. And even for patients who get on biologics, many of them end up cycling across therapies because they don't get remission with the first biologic they try. So that's a pretty good situation, I think, for us to enter into with the first-ever IL-23p19 in mirikizumab approved for ulcerative colitis. And we don't have head-to-head data in this indication, but certainly, the numeric efficacy numbers we saw were as good or better than any biologic in this disease state. So I think your assumption that IL-23s are going to be important for UC and IBD more broadly is a good one. I expect mirikizumab to do very well in IBD, that there aren't really small drugs in this category, a lot of growth possible given the number of people affected and the lower penetrants and the multiple lines of therapy that are offered.

Did you think TL1A data was differentiated over the IL-23 data?

I think it's hard to say comparing a Phase II to a Phase II trial, they seem roughly on par. I think the interesting thing about that mechanism of action is the ability to select potentially a subpopulation that can respond better so that biomarker-positive data is interesting if that's reproduced and if the 2 companies that have those 2 assets decide to pursue that as their indication. I think on the all-comers, probably not differentiated.

Got it. Okay. On the DICE molecule, it looks like it's active at the 200-milligram BID dose. But to really hit the target properly, at least on a patient-level scanner plot, 800 BID seems to be a right dose, which made me wonder, is 800 BID the viable program? Or should we really be thinking about the preclinical one entrant Phase I more potent?

Yes. 800 BID is a pretty high dose. We don't see a lot of drugs with that kind of dose. Of course, it all depends on efficacy and safety, right? Otherwise, this is just a number. The 200 mg dose looked like there was efficacy. I think it was a little bit hindered by a couple of patients who had really low PASI scores at baseline. There wasn't much opportunity for improvement and maybe another patient with discontinued therapy. So there's -- if we focus on the patients who had recently high PASI scores at baseline and stayed on drug, there's -- I think to my eyes, although it's small numbers, there's definite efficacy at 200, is it going to be enough to be competitive with biologics, which is a profile we want for an oral. I don't know yet. I think that's why we're doing the Phase II. So we're excited to see the Phase II for the lead compound. And then the next-gen compound looks quite a bit more potent. So my expectation is we could see similar efficacy to the 800 BID much larger.

Got it. So the deal was not exclusively on the backup molecules?

We purchased the company. We have a great team there and a great platform. I expect there to be multiple molecules coming from in the DICE group. In San Francisco, they remain intact and growing. And we're helping them develop more small molecules for IL-17 as well as other targets in immunology.

Got it. So speaking of M&A, I did want to ask 2 last ones. One is on POINT. I heard yesterday, I wasn't keeping close track of this, I heard yesterday that POINT is trading above where the bid came in.

Yes.

Can you -- what can you say, POINT overall and...

It's an unusual situation, a little bit confusing to us. I think Wall Street may be misreading some of the signals that they're getting here. Of course, POINT is a great pharmaceutical company working on lutetium PSMA, a drug which we think has great potential. It's partnered with Lantheus, where a lot of the economics will go. But Novartis released their data from PSMA 4, and I think there was some excitement about POINT, which is going to have an upcoming readout in SPLASH, thinking that the Novartis data created an opening for SPLASH to do even better. I'm not sure about that.

But you haven't seen the SPLASH yet? I think there's been...

No, we've not seen the data yet. And my baseline expectation is that there's no reason to expect it to be different than Novartis data. I think over-indexing on immature OS data is a bad idea. And I'm not sure it really matters either actually because the OS data that matter probably are going to be subsequent cuts of the data when it's more mature. Our interesting point is a lot around the platform, again, the ability to make multiple radiopharmaceuticals, I think is an interesting area of medicine. I personally have been in radiopharmaceuticals for a few decades now, and I think the time is right to get into therapeutic radiopharmaceuticals. They have a manufacturing facility under construction. And I hope we can complete the acquisition and help them build it, get it inspected and start launching products.

Got it. And just as we wrap it up, Dan, your overall thoughts on biz dev in general. It's interesting that you're being a little more active now but being very careful. There's an old big dollar checks. How are you thinking about that overall?

I think it's hard for any pharmaceutical company to generate value for investors if they're the high bidder in an auction of a derisked asset. So to say we'll never be the winner...

That's most of the large pharma space.

I know. But it really is converting that balance sheet to cash flow, which could be in a value-destroying way. I would rather be the acquirer of underappreciated assets that -- or maybe not easily understood. And that almost by definition means we'll be taking a minority view if we're going to win it. We're not taking the majority view and outbidding everyone else, that we see value where others don't or we have a unique way of finding value because of our portfolio, because of the strengths that we can add. It also means we'll go a little bit earlier than most. We have a bias towards great teams of people different than other companies. Some companies have a bias towards great assets, and then they dissolve the team and find cost synergies. Our play is find great teams and retain them and grow them and see what they can accomplish. That's played out pretty well for us so far.

Excellent. Joe, did we miss anything? Joe said thumbs up, so I think we're good to go.

Okay. No, thanks for the great questions.

Excellent. Great seeing you.

Great discussion today.

Absolutely.