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

[indiscernible].

Yes. However, that said, I'm sure you can put yourself in the FDA's shoes and realize that they don't approve a concept or a field or an overall direction. They have to approve a specific drug. And when it comes down to it, I would say that is going to be the predominant factor. But I think all of the data, I don't think it will just be ours. I think if they're getting -- looking for reassurance on the overall mechanism, I think it's the trajectory of the science rather than a specific study from that.

Let me ask you the other thing. I'm sure you watch data with us. Was there anything that you took away from the data presented or described at the document that made you change how you think about the design of your own pivotal? So one of the advantages you have -- you're the first pivotal program to be designed with the benefit of the aducanumab results. What did it taught you? What did the full results taught you that you actually want to use going forward?

I don't -- I didn't see something that was a make-or-break learning there. I think there's a couple things that we as a field, you really -- maybe I'm stating the obvious, but you don't want to start -- you do everything you can to make sure that you have no regret dosing, that you factored in as much of the future of the field as possible so that -- one of the things I think that was -- that everyone was in pain about was this whole question of was things happening different after they made that major amendment, was the placebo group somehow different, was the efficacy somehow different. And I think that that's something that if you can sort of see where the field is going and make the best guess as you can. But you really want to start these trials particularly with this -- within a field where so many things can be tough to monitor. No regret. And then the other thing is we also realize that tau PET is complicated. It makes it a little more difficult to start up new sites in new countries. But the fact that they did have some, that indicated that there was a hint of efficacy but not enough to really convince the committee. I think we will continue to make sure that biomarkers like tau PET are a major component of the trial. And that's just a component, yes.

Okay. So let's just go right there. So you guys have been, in addition -- well, you come from that world. But the tau PET imaging, where are we in terms of the frontiers of this? What do you guys have today that you didn't have a few years ago? And if you could just talk a little bit about how is this a different -- can this be used as a product? How should we think about the tracer world as something that we should control?

I think that it's -- it is a continually quite evolving field. It was, I think, 2012 when Amyvid was the first amyloid PET tracer. And earlier this year, end of May, Tauvid is now the tau PET tracer that's been approved. What we are doing is we want to make that available. Now we have -- all of the PET tracers for Alzheimer's disease are facing the difficulty of reimbursement because Medicare has not made a positive decision to reimburse. And they have had previous things on the books that say they will not reimburse new PET tracers for the brain. So they have to make an active decision to reimburse. There's no default to being reimbursed. So that is something that slows the field down. So on one hand, we have the advantage of now having 2 types of PET tracers: one that is able to rule out Alzheimer's disease; and the other, which is able to rule in the presence of the tau. That is one of the important 2 pathologies of Alzheimer's disease. So I think that begins a more sophisticated and more complete diagnostic ecosystem for companies like ourselves that see a disease-modifying treatment in the future. We are going to keep working on that. And hopefully, we'll get there sooner than later. But we don't want to forget that blood tests are also -- and over the summer, we had a really exciting paper in JAMA, where we have what's called a P-tau217 blood test. And I think we'll continue to expand as we sort of work with partners how that can be expanded into a blood test. We're going to look into that very, very carefully and do the necessary work. But if -- when that's -- if that's available, that will also dramatically make it easier for people to evaluate pathology.

So I think that's a beautiful segue into the next question. So I'm going to ask you -- because we're running out of time, I'm going to ask you 2 very quickly. So in TRAILBLAZER 2, you have added a blood test to this. Is this going to be enough to get it approved? And second, as a business of its own, because I don't think it's in our model or anyone else's model, how do we think about tracers as a financial contributor? Or are you thinking about this more as facilitating treatment as opposed to a kind of a business on its own?

The P-tau217 and TRAILBLAZER 2 is -- we see this as a way of doing this in a clinical trial setting to more rapidly test patients because of the beautiful correspondence. It's not 100% but it's high enough that we can use the P-tau test as a way of screening patients without having to wait for access to the PET scanner for amyloid and tau PET scanning. That's a really good plus as far as saving money and being able to recruit trials faster. And it gives us a lot of data for potentially looking down the road and helping people understand the diagnostic potential of these type of tests. So that's there. The economic -- your second question is a little more complicated because the economics of something as -- with a most incredibly complicated supply chain as PET tracers where it has a 2-hour half-life and so there's no inventory. You have to get up every morning and make anything that you're going to -- any doses you're going to have that day, you have to make it that morning. The supply chain is complicated. The economics are really complicated. There are, I think, definitely ways that it can add to the bottom line of any company, but it's going to take a process which probably will involve a reconsideration by large payers like Medicare before we get there. I think a disease-modifying therapy would certainly accelerate that reconsideration process. And so that is something that could change the economics a lot.

The last one I'm going to ask you before switching over and talking a little bit about diabetes is this issue of how to target tau itself. We talked about amyloid beta so far targeting. And what I was always playing with this is we don't really know how this -- I don't know how tau really participates in this process. It seems to be there. No genetic linkage but the biomarkers are there. And the one thing I had is this notion of why not use antisense as a way to knock down tau. It's kind of the more -- most elegant pan-tau approach to demonstrate that you're able to get the results as opposed to using like a Morphomer or antibody or something else. You always have pharmacokinetic questions, but I get into the right dose and so forth. So can you just talk to us a little bit about kind of the positives or negatives of using like an antisense approach or -- versus the small molecule or the antibody approach? Where are you standing on the antisense approach?

It's a great question. And you really -- you hit the entire spectrum of our multipronged attack on tau. We have the antibody that's actually in a Phase II that we think is going to have a very definitive test. Our antibody zagotenemab is, we think, highly selective for the tau aggregate. So a very specific species of tau that we hope that the fact that it does not go inside the neuron can still work because it's capturing those potential seeds as they go from one neuron to another. It's a very limited mechanism of action for stopping tauopathy. So if it works, fantastic, but I don't think it tests whether -- tau as a target. You bring up the other 2 really, really important ways of looking at this is can you interfere with the tau aggregation with something like a small molecule that goes in and can freely cross the blood brain barrier, goes right into the neurons that already have perhaps some tau deposits already in them and start reversing -- stopping or reversing them. So a small molecule could potentially have a huge amount of benefit. But we also know from many, many different research studies that tau aggregates across different models, even different mouse models, even different tauopathies in humans, takes on a really complicated variety of shapes. And so we may not be able to get there with a single small molecule that interferes with a tauopathy. But -- so that leaves us with the final thing is can we dramatically reduce the amount of tau made to slow that tauopathy. And that's where siRNAs and other types of variations on gene therapy could have a huge role. And I agree with you, I think it actually could be substantial. There's the downside, of course, that siRNAs do not -- at this point, negative siRNAs do not cross the blood brain barrier. So the ability to get them to neurons that are -- that have -- or are susceptible or already have tauopathy is going to be really tricky. I think we've shown that ASOs and siRNAs in certain different laboratory models the field has is what I'm saying can get into neurons and can effectively stop a target. I'm very excited about that potential for the field. And we have, as Andrew pointed out, our partner in Dicerna, and we are very excited about exploring this in a way that can give us brand-new ways of attacking the problem perhaps at the root. A lot of questions still left on that but very exciting direction.

So Mark, with that, thank you. And let me work -- move over to Jeff. Jeff, I kind of -- we got to you last, but you're probably dealing with the biggest questions of the day. Given that you guys got the 20th presentation to the investors, I'm not going to spend the next 15 minutes asking you about the tirzepatide readout. But the question I cannot avoid, there got to be 1 or 2, right? The question is, one, what is like a good outcome -- decent outcome and poor outcome in terms of a side effect profile to tirzepatide in your mind? Not just in the first trial but across the SURPASS program.

Yes. Thanks, Ronny. So listen, our objective for the Phase III program is to really replicate the kind of A1c reduction and weight loss we saw in the Phase II trials with tolerability similar to the pure GLP-1 receptor agonist. The results from the slower dose escalation in our Phase II dosing study we do think supports our expectations and reinforces the confidence that we have in our Phase III program. I know that many probably listening in today know about the dose escalation that we're using in the Phase III program. We use that Phase II data and modeling to inform a slow stepwise dose escalation. That's 2.5 milligrams every 4 weeks. And we do believe that will continue to then deliver the expected efficacy but also achieving an acceptable tolerability profile with lower discontinuations. I would also point to the -- even the 10-milligram data that we had from Phase II. The tolerability in that was very, very similar to what we see with Trulicity. And then another point I would make is similar extended dose escalation treatment algorithms have been used across the pure GLP-1 receptor agonist. One example I would point to is our own AWARD-11 study of the 3 and 4.5 milligram doses. In that study, we actually started all patients on 0.75 milligrams of Trulicity, escalated them to 1.5, 3 and then 4.5, depending on what dose arm they were at in at 4-week intervals. And that's actually different than what we employed in our original AWARD program for Trulicity, where we either just assigned them to 0.75 or 1.5. And in fact, the tolerability in terms of GI side effects that we saw at the higher doses in the AWARD-11 program is actually very comparable to what we've seen for 1.5 in prior AWARD studies. And I would say we've seen similar results in the STEP program with Novo for semaglutide where slow step-wise escalation certainly improves the tolerability for an incretin therapy. So we remain confident that -- in the dosing algorithm that we've employed. But as you pointed out, we're weeks away from our first readout of SURPASS 1. And at the end of the day, the data will tell the story.

You're going to get Christmas off or you're going to have to work with the holidays?

I don't know that we ever get Christmas completely off. We're hoping it will settle down but I -- we'll go through this a little bit on Friday. But I think everyone knows we didn't have a series of readouts kind of late first quarter, early second quarter of next year as well.

That's good. The other question I -- it's pretty obvious you're going to get some reduction of side effects. I guess the question is, are you setting us up for the -- are you setting us -- you're setting such a pretty high bar here, arguing you can bring it as low as Trulicity is in AWARD-11 just because you're still giving a very potent drug here. So you kind of told me, okay, we're going to go from 30, 40 to the low 20s. I'll go, okay. Yes, we can do that. But if you're talking about going to like nausea, vomiting and diarrhea rates, which are kind of like in a low-teens or single digit, that's a very high bar. I mean is this really as far as new models take you?

So again, I mean even for the GLP therapies across doses, you see it, right? And I'm not going to -- I wouldn't throw out a specific number per se. But I -- what we're saying is within the range of what's observed for pure GLP-1 receptor agonist. The other -- mechanistically, there's reason to believe that the GLP effect could actually modify the nausea and the vomiting. There's a lot of mechanistic data...

It would be helpful.

Yes. There's a lot of mechanistic data that suggests that. So again, our goal is to be within the range that you see with GLP therapies, realizing that there is a range even among GLP therapies. The other thing that I would point out that's important is we talk about clinical studies and comparisons to how drugs are used in the real world. The setting is a bit artificial in these studies and that patients are assigned to 1 of the 3 maintenance doses or placebo, if there's a placebo or a comparator, not based on whether they need that dose. And then -- so you have patients assigned to 15 milligrams that are then titrated to 15 milligrams over the 20 weeks, but it's not a treat to target study. And we know that in real life, there will be a lot of patients that perhaps are well controlled first at 5 milligrams, maybe for a significantly long period of time, then might later titrate to 10 and then later to 15. The scenario where everybody is having to titrate to 15 in 20 weeks is just probably not likely. That's one difference between, I guess, clinical studies. And then for drugs that have multiple doses, it's not different than an antihypertensive. I'll take myself. I take lisinopril. And I'm on what -- I think my physician called a whiff of lisinopril. I'm on a low dose. And you do that to make sure the person can tolerate it, they don't develop hypertension. But I fully expect that someday, I'll have to escalate my dose and it's a slow escalation. So no different here.

So I think what you're trying to say more broadly is we always think about clinical trials as being the best result and real-life results, when it comes to side effect, are often lower because the situation [indiscernible] control. While in this case, you're saying, no, no, no. We are forcing everybody to the highest -- to the target doses. In real life, what's probably going to happen is a lot of patients might have a multiyear titration or never even get to the highest dose. And the side effect profile might actually be better in real-life use just given the way the trial is being done.

Yes. Perhaps, perhaps.

All right. So the other side effect people have -- we have seen in this trial in the Phase II was hypoglycemia. And hypoglycemia became an issue and it was kind of pushing the Tresiba case against Lantus. So we all got exposed to that a little bit. How should we think about the hypoglycemia in the tirzepatide context? In terms of when it occurs, is that a risk, is something PCPs will be worried about? When you look at your drug versus others, what should we expect here?

Yes. So thanks for the question, Ronny. It's an interesting one that -- because we know that one of our competitor has also been making some noise around the hypoglycemia in our Phase II studies. Listen, the field is moving. And in fact, Novo has been a leader in moving the field to defining hypoglycemias, plasma glucose cutoff less than 54 or 52. When we look at the threshold of less than 54 in that Phase II dosing study that was presented at ADA last year, there were no patients that met the definition of less than 54 milligrams per deciliter. And when we look at our original larger Phase IIb study that was presented at EASD 2 years ago, there were very few patients, and they were generally balanced across the dose groups. Importantly, there were no cases of severe hypoglycemia in either of the studies. So I think when you look at it that way, now what was reported on the poster at ADA was the kind of the old definition of less than 70. But when you look at the kind of more contemporary definition of less than 54, we saw very little. Now you referenced also a comparison to insulin in your question. And it's, I guess, worthwhile to spend a minute on kind of the mechanistic differences here particularly as we think about hypoglycemia for insulins versus incretins. And that's with incretins, there's a plasma glucose dependency of the glucose-lowering effect. So from a mechanistic perspective, one could actually expect that tirzepatide could exhibit even lower hypoglycemia rates than a pure GLP because of the stimulation of glucagon secretion that occurs with GIP. It is -- GIP along with GLP, those are the 2 primary gut incretin hormones in humans. But in general, hypoglycemia risk with any incretin are typically only increased when they're used in combination with a sulfonylurea and/or an insulin. So we will certainly continue to monitor this as we see clinical data. But mechanistically, we wouldn't expect much. And then from the Phase II studies, when you look at that 54 definition, there just wasn't much there.

So when you say severe hypoglycemia, I just want to make sure we think about the same thing. This is a hypoglycemia which requires medical attention or health [ maintenance ]?

Yes. Yes. Correct.

But there were no cases where a person -- a measurement is a measurement, but there was never a person or a situation of a person who needed medical assistance as a result of hypoglycemia in your trial.

That's correct.

Is hypoglycemia something that happens early and you kind of like [indiscernible] in your trial? Or is this kind of like an ongoing situation?

Well, again, the numbers were so small in both of the Phase II trials that you really couldn't even draw a conclusion as to whether there was any time courses. We don't really believe there's truly a significant signal there. But typically, as you think about insulin, certainly during titration, hypoglycemia can be an issue. But then with illnesses, anxiety, stress, there could be other life happenings. And in the case of incretin therapies when used in combination with insulins or sulfonylureas, then you've got that risk that goes with those other drugs and then life circumstances that can alter that risk over the course of therapy.

So basically, there should be a warning on those drugs that if you are going to watch amyloid beta trial results, you probably should not be on them at least on a combination of those drugs. I see. Okay. So switching over and asking a little bit about obesity. I'm actually really excited about those drugs for obesity. And I'm kind of -- and Novo kind of went ahead and began a cardiovascular outcome trial for Ozempic. I was kind of wondering -- well, another dose of sema. I'm kind of wondering where you are in your thinking. Are you using the same doses of tirzepatide for obesity? Are we going to do cardiovascular outcome trials? Where -- how does you -- how far are you in terms of thinking about your obesity program? When can we see a potential approval? Kind of like this would be a really big market. So we're all very excited about seeing you guys move ahead in that direction.

Yes. Thanks for the question. We're excited about it as well. I would say it's kind of unique particularly outside of, say, our immunology space at Lilly to embark on multiple Phase III or Phase II programs for different indications, all in parallel with the primary indication. I know my first time in my 22 years, we've done something like that. So we're very excited about our obesity program. Let me just talk a minute about the program itself. So SURMOUNT 1, which is our longest and our largest obesity trial, we started it first because it is critical path because of the length. That was started in 2019. And actually, we completed enrollment of the SURMOUNT 1 trial in October. We had a brief pause in enrollment for COVID, but the trial was actually far ahead of our enrollment projections and still enrolled on time. So we really saw very little impact of COVID even on the enrollment in that trial. That study is evaluating the same doses as our type 2 program. We employ the same dose escalation scheme starting at 2.5. So yes, we are studying the same range of doses. At the end of the day, the Phase III trials will dictate what doses we take forward and request for approval. But right now, we're assuming in diabetes the 5, 10 and 15. And of course, we'll see what the obesity data look like as well. Beyond SURMOUNT 1, we do plan to initiate 3 additional SURMOUNT studies next year. And they'll evaluate weight loss in a broad population of patients, patients with and without diabetes. The SURMOUNT 1 study is in patients without diabetes. We also will include within that studies that look at maximizing weight loss and maintaining weight loss. So actually, we've said previously it's actually very similar to the STEP program. Now you're correct, we do not currently have in our clinical plan a separate cardiovascular outcomes trial in patients living with obesity and without diabetes. That's not a distinct requirement for FDA and we've aligned the program with FDA. Now having said that, it doesn't mean we don't -- aren't interested in outcomes. And in fact, we're very interested in outcomes as a part of our research strategy, which is tirzepatide. So we initiated the SURPASS cardiovascular outcome trial earlier this year back in June, which is unique in this arena. It will be the first head-to-head cardiovascular outcome trial, powered for both noninferiority and then superiority Trulicity. We feel that's important because I think that's going to be the next question that prescribers have is, okay, we know the GLPs so far have generally demonstrated cardiovascular benefit. Do the next-generation therapies have the same benefit or even more? And we do have reason to think with some of the markers we've seen in terms of triglycerides, lipoproteins, cardiovascular inflammatory markers and then just the weight loss that tirzepatide could even be better than Trulicity in terms of cardiovascular outcomes. We also have our NASH Phase II study ongoing, as you probably know. And assuming that yields positive results, we're talking about a Phase III study in NASH, which would incorporate outcomes. And then finally, I would say we continue to discuss other potential complications related to obesity that we could pursue in order to demonstrate additional benefits in that patient population. So more perhaps to come on that later. Right now, our focus is on seeing some of the initial SURPASS data.

I hear you. So not committing yet to a separate cardiovascular outcome in obesity. I mean the argument there simply is that if you want to eventually convince CMS to start reimbursing this, right, it will be good to see CV data. But I'll let you guys finish readout of the trials and ask you the same question 6 months. How's that?

Yes. That's fine.

And last but not least question just because we're running off time, NASH. So we have some interesting data over the weekend. And the question for you is, look -- I mean 2 questions. One of them is you kind of look at the trials being done in NASH. And I call them being tested in near-unicorn, right? They're testing people who have controlled hypertension, controlled obesity, controlled diabetes and still have NASH, which obviously is not the NASH population. I guess there's a question about GLP-1 just because of the mechanism being able to kind of do the trial in real patients as opposed to made-up population. So if you could comment on that. And second, how are you thinking about the combination of -- clearly, the GLP-1 works, but a combination of GLP-1 with other modalities? Is this the right way to start? Or look, GLP-1 is so efficacious on its own, it's kind of nice to think about it being combined with something else but let's just get those through the pipeline first and then worry about -- let them worry about how they will be recombined? How are you thinking about that?

Yes. Big question. And you're right, we've -- we're seeing increasing amounts of data at least on the GLPs all the time. You're correct, firstly, that there is significant overlap between the disease states that we're studying obesity, diabetes, NASH. But the F2, F3 population, which is the focus of our study right now, is a large heterogeneous group, and not all of them even have controlled diabetes or controlled -- or have the urgency to seek a treatment for obesity right now, particularly when available therapies result in, what, 5% to 8% body weight loss. But we're looking at therapies that in the future -- and even in our Phase II diabetes study that wasn't a selected obesity population, we saw a 15% weight loss at the higher doses for tirzepatide. So I think -- and given the associations with weight, fatty liver, et cetera, there's great opportunity for next-generation therapies that can really impact weight in those patients with obesity and NASH. I would also say drivers of therapy in obesity are not just the BMI but having comorbidities. So having an indication for NASH for someone who's obese, maybe they would not choose to take a GLP or maybe in the future, potentially tirzepatide. But knowing that they have NASH and that additional risk factor could drive them to the therapy. And then we know there's significant overlap between NASH, obviously, not only the liver disease and liver outcomes that come with obesity but also MACE events. So you look at this constellation of potential negative outcomes for patients in this overlapping group of what you called the near-unicorn, but this overlapping group of patients, and we still think there's an awful lot of opportunity there. Now most of your question was focused on what we know about pure GLP, GLP-1 receptor agonist. But I think as we look to next-generation incretins like our dual agonist tirzepatide and we think about some of the unique pharmacology, particularly the GIP component, that's where we get very excited, I mean the potential to impact ectopic fat accumulation in the liver. We've got a weight-independent effect on insulin sensitivity. And we reported out some of the biomarker data from our Phase II study that wasn't even a selected NASH population with pretty impactful results on those biomarkers. So ultimately, I think our belief is that the dual agonist could take us somewhere in NASH that a pure GLP agonist would not. We're keeping an eye on the combination space as well. And I think you know we've got other assets in development, both incretins and non-incretins and both in our own portfolio. And we're always looking at business development opportunities as well. So we have not ruled out a combination approach but we've got to first generate some data on tirzepatide alone.

Very good. So with that, given that we are right sitting there at 12:00, I want to thank the Lilly team for being with us today. And hopefully, we'll get to do this again sometime next year. Thank you and have a great day, everybody.

Thanks.

Thank you.

Bye-bye.

Bye-bye.