Merck & Co., Inc. (MRK) Earnings Call Transcript & Summary

Hi, everyone. Thank you for coming to the session. My name is Daina Graybosch. I and my team cover immuno-oncology and beyond here at Leerink Partners. And I'm really excited to have a great oncology conversation with Merck. We have Dr. Marjorie Green and Peter Dannenbaum here. And I thought maybe we'd start. Dr. Green, you could give a bit of an intro to yourself because you're relatively new to Merck, and then we'll jump into some questions on oncology and Merck.

Fantastic. Thanks, Daina. Really excited to be here and for the opportunity. So I have joined Merck in April of 2023. By training, I am a breast medical oncologist, so I'm a sub-subspecialist and was at MD Anderson for a long time on faculty there and then moved to industry and have had the pleasure to work at Genentech and then most recently at Seagen prior to joining Merck. Since joining Merck, we've had continued expansion of the portfolio, and I'm now overseeing both early and late-stage oncology. And I'm really excited about the future for patients, given the strong history of excellence, the foundation of KEYTRUDA and then this very diversified, fantastic portfolio that's been built.

And so how we spend our 28 minutes, which is not nearly enough time to talk about Merck oncology. We're going to start with some ADCs, then KEYTRUDA, going to talk about the vaccine and then sort of novel IO targets towards the end. And maybe we'll leave 5 minutes if people in the room have a question as well.

So let's talk about ADCs because that's relatively somewhat new for Merck, a little bit of programs for a while. What's Merck's strategic approach? And also, I wonder how you think today, where Merck is coming in, how -- what's a successful ADC strategy, and how does that differ from strategies people employed, companies employed 5 years ago?

Thanks for the question. Merck has been very deliberate in how we have built our oncology portfolio. Historically, chemotherapy has been the foundation of care for people with cancer, multiple different kinds of malignancy. And with KEYTRUDA, that has moved to really change the paradigm. And so people used to say, what do we add on to chemotherapy? What can we do with chemotherapy? Now it's how do we work with KEYTRUDA? In many of the malignancies. And so if you think about what ADCs are, they are more potent targeted chemotherapeutics usually. They deliver a cytotoxic payload, and so they are often able to be more active than the parent or comparable chemotherapeutics. So for an example, [ tecan ] is there is a cousin to the topo1 payloads, and breast cancer had modest activity. But when you look at medicines like in HER2, you're looking at 2870, the data is much more compelling. So they're very potent cytotoxics. And they're also very accessible, too, in that they are drugs that can be given in an academic center, that can be given in a community center. And clinicians know how to manage chemotherapy. And so thinking about where you want to go with medicines for patients, it's always based upon the unmet need. But the science of the ADCs has made it to where there's better targeting of tissue, the cancer tissue, with less off-target sort of side effects from the ADCs. So from a strategic standpoint, we look at it and say, chemotherapy and KEYTRUDA is core to many of our disease areas where we've invested and where patients have unmet need, so let's see what ADCs can bring to this for patients. So it made a lot of sense for us strategically to move into ADCs. I think that there's a tremendous amount of advances that are being made continuously in ADCs. You see the number of companies that are investing and building in ADCs. And that leads -- it's because of the excitement and the potential of what's there. ADCs are so complicated and what they do in that they have the antigen that you're targeting, the antibody, the linker, the payload. All of those are designed incredibly carefully to optimize the delivery of the cytotoxic into cancer cells. And there have been multiple learnings over the past 20-something years that ADCs have been in development. And there really have been some technological breakthroughs that have made these molecules much more accessible for development and more impactful for patients. So it makes strategic sense for us to be here. And thinking about the future and what we'll be doing over the next 5 years, I think that we're in an area where precision medicine, where the ability to identify a patient and the tumor that they have. And whole benefits becoming increasingly much the norm in oncology care, it's a need and it's a norm. So for example, in lung cancer, EGFR mutations, ROS, ALK, other mutations are being identified, and then treatments are being posed for it. Breast cancer, whether it's estrogen receptor status or HER2 status, multiple tumor types, you're seeing this. And for ADCs, they lend themselves to wanting to have more precision in how they're used. So whether that is, for some ADCs, if the antigen has lower expression, you definitely need a biomarker to do that. For others, there might be different approaches, either the strategy and how you move forward in combination or building on IO therapy or small molecules. And our 5-year sort of approach, how we're going to do it, is first, we have a great history of taking a medication and then adapting a diagnostic to it, so with KEYTRUDA. And so we've got the knowledge and the skill set of being able to think about ADC development with that lens, where it's needed, that we'll be able to do it, and we'll do the investment and the science to figure that out. Combinations. We have built a very robust portfolio of multiple drugs that lend themselves to combinations with ADCs. And I think it gives us very much a strategic advantage that will really benefit patients based very strongly in science and rational combinations. And then we have KEYTRUDA. So people keep talking about the LOE for KEYTRUDA. But KEYTRUDA gives us so much information and is complementary with ADCs in a similar way that it is with chemotherapy. So our strategy moving forward is going to be based upon those kind of core ways of thinking about ADC development. We also have extensive discovery work trying to improve upon technology. So I think you'll see a lot of advances there as well.

On the technology side, do you see Merck as an acquirer or innovator or both with sort of novel designs for ADCs?

I think both. And it's because one company, it's hard for us to be perfect and do everything. And so we want to be open and opportunistic when needed. And so you'll see that with some of the work we've done with Daiichi and with Kelun, is that these agreements that we have with both of these fantastic companies, the drugs are different, the targets are different. And so we know that they'll benefit patients and they're scientifically very sound. Our discovery group is making advances in ADCs, but we can't do everything. And so where we think the science is better or stronger or different than what we're doing, we will likely be able to look externally as well.

You mentioned biomarker in your history making, I guess we could argue, the best companion diagnostic for KEYTRUDA. How are you thinking about it across your portfolio? So TROP2, HER3, CDH6, B7-H3. Will you be looking at potential companion diagnostic always for all of them? Or are you already narrowing in for, some of those will need it, and others will not?

I think it's -- it will probably be some will need it, and some will not. And the baseline levels of expression of an antigen as well as the sensitivity of the payload really lend itself to understanding how much do you need to enrich versus not. Most of the ADCs developed today have bystander effect, where because of heterogeneity of expression of the antigen target in tumors, by having an antibody that can kill cancer cells around it that don't actually have the expression, you don't always need to have expression to have activity. But the bar is getting higher and higher. So you think about KEYNOTE-189 as an example in first-line nonsmall cell lung cancer. To be able to replace the chemotherapy portion of that, it's a really high bar. So enrichment might make sense in that kind of situation if combinations or other approaches don't work. So it's going to be nuanced and fit based upon the disease indication, the asset, the payload and sort of the clinical need. And so it isn't a plug-and-play formation. But I think it just -- like it wasn't -- with looking at our PD-L1 assays, it's not the same across all of the different disease indications. And I think we're very fortunate to have the resources and knowledge to be able to adapt because it won't be one size fits all.

On the KEYTRUDA combo, your third point, third leg of the strategy. I think it's pretty fair at this point to assert some real synergy with KEYTRUDA and ADCs in some situations. And I wonder how you think about it scientifically. Like sometimes, it doesn't look as synergistic to my eye, it's early, and sometimes it really does. Like what's driving that? And how does that guide how you think about combining with KEYTRUDA going forward?

For the question. From the basic, you think back to sort of some of the science behind IO combinations, there was always a question at first, if you give chemotherapy with a -- like KEYTRUDA, would you actually diminish the effect? But then the other school of thought was, well, with more antigen presentation, you actually will get combination data that will really be potent. And I think we see that just sort of baseline. It's much more like it -- the floor is that it should combine generally with ADC and an IO therapy. The question of synergy is tricky because how do you prove that clinically? So there are some companies, and I might have worked at one of them, who will talk about that. And scientifically, it's known that some chemotherapeutics do induce immunogenic cell death, and maybe you'll get some benefit from a combination there. And not all of them do. But most of the data that we see when you look at it, if you can get something that's at least additive, that might be what you're calling a synergy, is that usually a combined drug that's rarely it's A plus B equals really the numeric what it is. You get some kind of incremental improvement on that. And so it's -- there's -- how you select for it, I think, is complicated because there isn't a hallmark of, biologically, can you predict where synergy is going to exist? I think it's more complicated than some people have talked about it.

So you're not saying, hey, we will always try to go after a payload that has preclinical synergy IO or a certain FC. It's more like, look, we'll be broader, and we're going to see at least additive effects. It's not just we want to have just synergy...

I don't know -- I have to be careful again about my former employment obligations. But I think synergy, it's hard to prove clinically. And so I want to see the efficacy that you consider to be transformative. That's always our goal. And so if we can do that, either through how we combine, how we select patients, that's always our goal. And so it's not necessarily the combination with KEYTRUDA that's -- the payload and the ADC that's always going to determine it, but it's part of it. And so we have to look at all of those different opportunities to get to that level of efficacy that's really transformative.

So maybe some quick-fire. I said quick-fire, let's see. This first one is a short question, but I think I'm asking a broader Merck question at the same time. So you have an IDH2 study. That's for your B7-H3 ADC with Daiichi. It's a Phase III with dual ORR and OS end points. And I wonder if we -- can we infer that you might use the ORR to file for accelerated approval? And I guess the broader question is with FrontRunner and as you're thinking about oncology development, is this going to typically be your accelerated approval strategy going forward, where you might use an interim and you might not?

Yes, no, thanks for the question. This is always tricky because I love talking about this stuff, but also this goes into the confidentiality of we rarely disclose what we're doing with our registration plans. We usually have more than one endpoint for efficacy in our studies, and there are different reasons. Sometimes it does get it into the label. Sometimes you could consider for an accelerated approval. There are lots of reasons. Sometimes it's because it's clinically meaningful to a population, and you need to demonstrate that for clinicians and payers. And so it's really always hard to tease that out, when we have our dual endpoints, what the goal is when we put them in there. But it's very normal for us to do. And we try to set up our strategies in our statistical designs where we're able to independently test endpoints. So I think the Project FrontRunner is fantastic. What it does is it gives you the opportunity to be able to potentially get accelerated approval in earlier lines based upon an ongoing Phase III study. So it derisks it somewhat for the FDA, but it also allows you to get access to patients earlier in earlier lines of disease. So I'm thrilled that they're doing it. And so we do always look at opportunities to consider it for our Phase III trials. But usually, our dual endpoints, it's sometimes, there are lots of reasons why they're in there.

Second one, a hot topic on TROP2-ADCs. You have a Phase III in -- with your TROP2-ADC now in frontline PD-L1-high lung cancer. And I wonder if you're enrolling both squamous and nonsquamous and why, what supported that decision?

No, thanks for the question. We're excited about 2870. It's definitely a differentiated ADC. It's -- when you've got similar targets, people keep thinking, well, why will we need another ADC? And they're not identical. They're not the same drug as we go from ADC to ADC. And we really like the -- what we're seeing with the Kelun asset as they are already in Phase III trials and some of those are reading out. The TROP2 expression is equivalent between squamous and nonsquamous. It doesn't really seem to be an expression difference between them. And we're building our studies based upon emerging data out of our ongoing Phase I, Phase II studies. And so we have enough confidence to be able to put both of those into the study, but we are stratifying. So we'll have an even balance between arms. And so we'll be able to -- in case there is a differential effect in a first-line combination setting because that's -- it's not a monotherapy play, it is in combination, that we are going to have those balanced and be able to look at that.

I wrote a long paragraph about TROP2 biomarkers which I'm not going to read because this is sort of my meandering hypotheses. But something seems to be going on with it, with you and competitors talking about it as a diagnostic, we haven't really seen a lot published. Is there something hard about TROP2 as a diagnostic? And I think is it fair to assume that you will be using that in some of your -- as an enrichment strategy in some trials?

We have -- so starting as this TROP2, part of the challenge goes -- first let's go back to the ADCs that have bystander effect. So when you have -- and also where do we get our tissue biopsy? So patients in metastatic disease in our studies, they often have their biopsy at some time ago from when they're actually being treated. And so you're going archival tissue and getting a sample of what you think an expression in a tumor might be. And there's a lot of heterogeneity in a tumor so -- throughout the body and the tumor itself. So the bystander effect is important. When you look at activity, you often see it over the range of IHC or H scores or whatever you want to call the whatever kind of diagnostic that you're using. So making a cutoff can sometimes be tricky because you don't like leaving patients and efficacy behind. If someone has the potential for the drug, it's very hard to not want to be able to optimize it. And there's been a lot of that done. I think even going back to public data from TRODELVY, when they looked at triple-negative breast cancer, they were able to enrich when they looked. It didn't mean that patients who had low levels of TROP2 didn't benefit, but there was a differential activity level based upon expression. So some of it has to do with what are you trying to accomplish with your study? And I think that there are probably going to be studies where we do enrich and we may look at high biomarker populations because either the comparative bar is so high or we want to ensure that we've got that transformative benefit. So it is we're building the airplane as we're flying it to some degree, I think all of us are who are doing ADC work. But we have the tools to be able to do it. So I think you see people talking about it because, when you look at the data is out there, it's good, but we could be doing better. And biomarkers might be one way to get there.

I'm going to move and talk about KEYTRUDA and subcu. Probably for me, the most interesting thing in the 4Q call came in the Q&A when Rob, in answer to a question, suggested you may be able to launch subcu KEYTRUDA in an auto-injector, meaning you could get it in the home use. I just wonder how likely is that outcome? And was there a special sauce to get the milligrams of KEYTRUDA low enough to enable an auto-injector?

I'm going to call on a friend here in a little bit with this one, but Peter is sitting next to me because he's got the longer history with the company than I do. The study that we have ongoing with our KEYTRUDA subcu is not given with an auto-injector, and so the study that we've got primary completion late this year. However, you can imagine the potential and the opportunity. Right now, checkpoint inhibitors are given in the clinic, in the hospital. Is there the potential, because we know the safety and the tolerability of KEYTRUDA, to be able to move to a near future where this is given at home? I think it's not unreasonable to consider that, particularly as many of our studies that are ongoing right now with KEYTRUDA are reading out, are in the curative setting. And so once chemotherapy has finished, there's a lot of potential for the ability to administer medications, either very, very quickly with an auto-injector, or the subcu formulation has a lot of value to patients. I think that we're not ready for the time that Phase III reads out for an auto-injector. So I think I want to be clear about that. Beyond that, I don't know what Rob was thinking as he said it. So -- but curious, Peter, if you have additional information to add from your perspective.

So our subcu program combined with hyaluronidase, the coformulation, reads out this fall, as Marjorie said puts us on track if successful for an approval some time in early 2026. The auto-injector would come later. But we're very excited by the innovation that will be provided to patients with the subcu. We think it's needed in the space, particularly as you think about how extensive KEYTRUDA is becoming in earlier-stage cancer patients.

Yes. Think about nonsmall cell lung cancer, triple-negative, receptor-negative breast cancer. No one wants to have [ por ten ] and be burdened by the IV therapy. So I think there's a lot of patient advantage and innovation by doing this.

Maybe let's move to some other innovative programs, starting with the program, the INT program in collaboration with Moderna. For me, I still would really like to see more biologic correlative data with the immunogenicity and clinical response. And I wonder what for you guys, other than the Phase II findings on RFS, DMFS, which was randomized, gives you confidence that INT is contributing to benefit -- enough confidence that you've started additional Phase III trials?

Thanks for the question. It's -- I can understand definitely from an external perspective wanting to see additional data. We've generated it. And so you'll -- what you'll see is the more information coming, more studies, Phase II studies, Phase II, Phase III starting up. And the first one hit, I think ct.gov this week or last week. We had a question about it in an earlier session in cutaneous squamous cell carcinoma. So there's -- Moderna has 2 studies ongoing right now where a significant amount of correlative data is being generated. We haven't disclosed any public plans on when we would share that information or any of the early data, but we do look at it. And it does inform for the work that we do. We also -- we keep an eye on what's going on in the competitor field. But you want to have some equipoise before you go into clinical study that you're actually adding benefit and value. And so we don't lightly go into making Phase II or Phase III investments without having confidence on the science and the potential clinical benefit that informs.

Of the KEYNOTE-942, it's been recent questions on censoring. And it appears, just looking at the censoring ticks and the active and control, that there's a bit of a misbalance with censoring early on in the active. I wonder if you could talk about, is that just chance? Is that a particular kind of informative censoring? And how we should interpret that data?

Yes. That study, Moderna operationalized the study and did a really, really great job with it and was recently published a month or 2 ago. So full details are available online. The study was being conducted during the height of COVID. And there is a period where Moderna had just shifts in their manufacturing capabilities into actual COVID vaccines. And so there was a period where there was actual assignment of treatment as opposed to full randomization. And so there's a period where you had people on KEYTRUDA only and not getting INT because of manufacturing capacity and the need to prioritize COVID vaccines. And this was early in 2021. And then once that went away, then the normal distribution happened. So you have a weird censoring pattern that appears in that 12- to 18-month period. And when we do the next update, it will probably shift out to the 24- to 28-month period. And it's just related to that need to be pragmatic during a pandemic while maintaining your balance between your treatment arms. I think Moderna did a really nice job handling that complicated issue that they had going on there. What I can tell you about the censoring, and this is in the manuscript, is that there is no censoring for starting new therapies. I mean, basically, people were censored because they're alive and well.

Very helpful. Looking at 5 minutes. I wonder -- I'll just ask a broad question. And we do see that there's some data coming at AACR. But as you think about the earlier-stage portfolio of oncology assets, which ones would you highlight as you been particularly excited about or that we would see data this year?

Seeing data this year, I can't comment on that directly. I'm newly over the early stage portfolio. And so I can't comment on which ones you're going to be seeing data from this year. I'm excited about definitely some of the ADCs that we have going into our early stage, so 1200, MK-1200, which is our claudin 18.2 ADC, is in the clinic. It's a Kelun collaboration ADC that's been brought in. That was exciting. We do have an early -- I'm sorry?

[indiscernible]

Claudin 18.2 is highly expressed in GI malignancies, so. Then we also have KRAS, our G12C inhibitor, is moving from early into late stage this year. And so we're excited about that. I think that one of the things that I love about it is that it has been dose -- it's been manufactured and optimized to really be therapeutically potent in a way that's an advance on the original G12C inhibitors, and it's very combinable. And think about our broad portfolio and what we have, it combines really well with KEYTRUDA. We're excited about the potential of that drug also.

I'll open it up to the audience. Any questions? Then I'll ask some. So TIGIT. What gives you confidence in vibostolimab and your development strategy? And then maybe would you ever and are you considering a biomarker strategy for TIGIT?

TIGIT is still in early days. It doesn't feel that way in some ways because we've been talking about TIGIT for a while, and there was this big push by Roche to get Phase IIIs, then everyone started following to do that. But we're still in infancy in TIGIT and learning about it. But what we can see -- say so far is that you're seeing activity. And so this is where we do have 5 Phase IIIs that are ongoing. And primary completion dates go out into 2025 and beyond. You'll start seeing data coming from them. And we have Phase IIs that are informing for us as well. We have our studies set up in a way that we can ask lots of questions. So like I talked about in the endpoints, so in one of our studies, we did add a TPS greater than 50 population. So we're able to look at the totality of the TPS 1% and higher as well as the 50% and higher just because there's a lot still that we don't know yet about TIGIT. I think that it's active. I think that it is something that adds on to KEYTRUDA. And as these studies read out, you'll see modifications and ways and potentially new kinds of studies that take advantage of the knowledge that we're gaining.

Let me ask about BTK. That's one that you guys acquired a long time ago from ArQule, nemtabrutinib. And you recently have started some new studies. And I wonder if you can remind us what's differentiating about that? And what took sort of -- took a while to get this going in late-stage development. And correct me if you think I'm wrong, but how are you thinking about what you can do with that, given where competition has moved?

So nemtabrutinib is -- it's a noncovalent BTK inhibitor that we acquired and brought in. And it's now ongoing in phase -- in 3 Phase IIIs. So once we got the big Merck ship going, it's going. And so we like the medicine because it is specifically designed not only to be very potent, like you see in some of the newer-generation BTK inhibitors, but also to work quite well where there are mutations presence, where there's the 4 something, something, [ 468S ] I think, is the mutation that exists. And it may also prevent the mutations from developing with time. So because of that, it's particularly attractive for earlier lines of therapy. And that's where you see some of our studies that are actually being conducted are in the first line, is because we think it's got the potential to provide even longer, more durable control for patients by being able to prevent these mutations that cause the resistance and the relapse compared to therapies that are already out on the market. So Peter, this is before my time. Can you speak at all to sort of the time to start?

It took us time to get the dose right. So we wanted to make sure we have the right dose before we develop further.

And what did you do? Was that additional clinical work or that was like preclinical work?

It's additional Phase II work that we did, yes.

It was clinical but there was -- they were a bunch of PK analyses, exactly. So thank you.

I wonder if you could talk about -- last question. No, go ahead, Peter.

Yes. So I just want to follow up with [ Peter. ] So how did you deal with that PD-L1 based on lung [indiscernible]? What's your [indiscernible] combination for PD-L1 based on [indiscernible] of the 2 years [indiscernible].

There are multiple studies in lung cancer that are looking across different levels of PD-L1 expression. We do know that generally, the higher the PD-L1, the stronger the benefit in nonsmall cell lung cancer. But you still see long-term overall survival advantage across the indication, which is why we've got broad indications with our studies. They were all designed a little differently, so sometimes it's tricky to tease out from one study to the next, the difference in there. But this is part of the -- as you go into earlier lines of therapy, the immune system is more robust, which is why you can see for, like an example, in breast cancer, Y522, you're able to have an all-comer label and benefit. Whereas in the metastatic setting in 355, it's greater than 10%. And so there is probably some heterogeneity in the population in the early lines of therapy, but chemotherapy does increase antigen expression. And we think that's why we're able to see benefit long term in these populations.

Okay. Well, that 0.5 hours went really fast. So thank you so much for your time.

Thank you very much. Thank you, everybody, for your attention.

Thanks so much.