Skip to main content

Merck & Co., Inc Investor Event Transcript, Nov 19, 2026

Merck & Co., Inc. (MRK)

Merck said the FDA wants trial data that include U.S. patients, after the Harmony CRL. Executives said the maturity of U.S. data matters more than the percentage of U.S. patients. They also described MK1084 as well tolerated. AI summary

Investor Event Transcript 2026-11-19 For: 2026-12-31
Added on November 19, 2026

Conference Transcript - MRK 2026-11-19

Akash Tiwari, Other

Day four of our Jeffries Healthcare Conference. This is not one of those conferences where you can leave on the second day. Which I love doing sometimes, but it's a long haul, and I really do appreciate everyone sticking around. I can tell you personally, it's been a joy to see all the clients and all the companies, and it does seem to be such an exciting time in health care. My name's Akash Tiwari. I am head of our biotech and pharma research efforts, and we have Merck and Eliev and a fireside that I'm really looking forward to. Why don't I hand it off to you for some intro remarks and we'll get started.

Speaker 0

Oh, great. Well, thanks. Thanks very much for the invite. I really appreciate it. And I'm really happy to be here at a time when we're able to see a really exciting R&D, a period of R&D for the company. We've had, over the past few years, the beginning of the transformation into the meaningful post-Ketruda era, with nine new medicines and vaccines being approved over the past few years. We've had some really exciting phase three readouts, including SAC-TMT and Antismoran, Autogene, as well as our HIV portfolio. And we're going to have a very rich, data-rich next few months, both in Phase III and in important Phase II proof-of-concept studies with drugs like alamatrivir, tulisakibar, and others. I think that the The strategy that we laid out of diversification and following the sciences really paid off. Not everything is going to work, because that's science, and science is hard. But I think overall, we're very optimistic that we'll be able to provide a very meaningful set of data that will help patients improve and extend their lives, help communities prevent and return good value to shareholders.

Akash Tiwari, Other

Love it. So I want to start off with cancer vaccines, and I know there's a lot of focus on the renal data that there may or may not be an interim. As Peter reminds me, Merck does not comment on interims. But I actually want to focus on something more important. I remember at ESMO, you know, your team's talked about, in general, like, hey, there's one of the dynamics here is TMB, high and low. But I feel like you've also learned a lot, and there is more nuance. It's beyond just TMB. There's IO sensitivity. It could be the types of mutations. It could be kind of neoantigen load itself. So there's definitely, I think, a more dynamic framework. of where a cancer vaccine works and where it doesn't. So... You know, your publication came out with Moderna. We were waiting for some more information. I know that's going to take some time. So maybe help us with what your evolving view is beyond just TMB and responsiveness to a cancer vaccine.

Speaker 0

Yeah, so this is all new biology, and it's a very, very exciting time. We're thrilled with the results of the INT-001 study in melanoma. We're learning a lot from it. It's a very data-rich program, and we're going to be able to make a lot of important discoveries is about what constitutes how efficacy might have been better or less well, you know, how we could improve algorithms, what are the relevant patient populations that particularly responded, and so on. I think the Phase II program that we have going forward also provides us with an important assessment of how wide is the aperture for INT. Thank you. Thank you. Do we have to include, is Keytruda in background necessary? Is it IO-responsive tumors that's important? Can you give the INT by itself? Can you give the INT with other immune stimulants like BCG? And so all of that work will provide a lot of the answers that you're asking about, but I think it won't be just TMB, that's for sure. It's going to be It's going to be HLA. It's going to be the tumor microenvironment. It's, you know, we're looking at, we have a couple of studies in the metastatic setting just to see how that looks. But, you know, again, it's new biology. We'll have to see how it goes.

Akash Tiwari, Other

Understood. And maybe just as a follow-up to that, your team did have interesting data in pancreatic, as did BioNTech. But I could also tell that that's not necessarily where you're steering investors. It's an interesting proof of concept. So talk about why renal. You feel like that was in that initial set of bets we're making, whereas for pancreatic, that really wasn't, you know, the highest probability bet for your team. Sure.

Speaker 0

Well, I think that pancreatic cancer is very cold immune from the immune response point of view. There was never much response with Pembro or any of the other immunotherapies. You know, it's a, the patients are fairly advanced when they get, when they get diagnosed. It's not, there's not very few patients that can undergo surgery any, when they do, it's a pretty gargantuan surgery. Whereas renal cell cancer was, is incredibly well studied. We've had multiple trials in the renal cancer setting, renal cell cancer, and in early. So, the combination of Pembro in early disease, immunoresponsive, was the first, was kind of the beachhead of where we were going to go. If we see strong results, we'll then expand even further and ask the questions about cold tumors, but that's not the first place to go. So even, you know, you see some phase one proof of concept. Right now, I think physicians who are interested in pancreatic and treating pancreatic cancer are focusing on the RAS inhibitors.

Akash Tiwari, Other

Understood. Now, you know, on the interim, which I know you're not going to comment on, I'm going to actually shift in a different way. I feel like there's too much focus on the interim when I saw the landmark analysis you showed in melanoma after a year. Where, you know, it's not that large of a study. You can see enough variability in terms of how I.O. can perform, but the real thing that stands out, at least to me when I look at your melanoma data, is it works in PD-1 high, it works in PD-1 low. We didn't see that with LAG-3, and it's the duration of response. On top of, obviously, the side effect profile. So let's skip the interim question. But why is your team interested in looking at the long-term follow-up in renal? And why would that maybe be the right focus for Merck?

Speaker 0

Yeah, so, I mean, you have to, so thank you, and I agree with you about, you know, the nature of the, you know, of course, interim analyses tell you different things. And, you know, I know investors really want to know as soon as possible where things are going, but it's, science has its own Rhythm and timing. If you look at the analysis of the melanoma data, you saw this piecewise analysis where the first few months there really wasn't much of a response. Then afterwards you had this huge response that continued and improved over time. That is completely consistent with the immunotherapy approach and why we use nine doses. We need to prime and then repeatedly And so the upside of that is, is that, you know, obviously you get long-term responses that make all of us here excited. The downside is, is that it's the enemy of interim analyses, because it's, because there's a, you know, there's a time basis that does not, that simply, there is no amount of money that you could throw at it that'll change, that it takes time. in terms of follow-up time. I'm not talking about time to recruit. So I think it's going to be important for us to not overreact to any interim analyses, yes, no, whatever, and to look at the overall picture. The renal data are very important because it's a TMB low patient population. So it tells us whether or not we need to focus just on the TMB high tumors, which by itself is a huge It's a huge opportunity, but it tells us, like, okay, you know, you really need to stay within the sweet spot of Keytruda and focus in that area. But, you know, again, I think doing the study at this stage was important because we needed to understand whether it was Keytruda responsiveness or TMB or both.

Akash Tiwari, Other

Understood. Now, one of the really provocative data points in that Phase II publication, but unfortunately I think it was only in 10% of your patients, was when you looked at longitudinal CTNA positivity and negativity, that was in of itself actually the biggest predictor of whether a patient was responsive or non-responsive to a cancer vaccine. Unfortunately, because it was such a small subset, we can't apply that to the full group, and I'm suspecting you might have more data. But let's say... That if you did measure that in every one of those patients in your phase two, and that signal held up, what conclusions could you make about whether TMB is driving or non-driving the responsiveness to a cancer vaccine?

Speaker 0

Well, I mean, I think, you know, ctDNA's role in monitoring patients' responses is a relatively new concept. I mean, ctDNA's been around for a long time, but as we build more and more ultra-sensitive tools, we get more information about that. The issue with ctDNA, to be frank, is that its shedding is very much tumor-dependent, different tumors of different shedding kinetics. When you have larger, there's a bias, an intrinsic bias about the size of the cancer and its location, just because of what gets, how much shedding happens. So I'm not so sure that that's going to be a great tool for assessing. Responsiveness or something like that. What I do think is going to be important, and I don't have any data right now because it's required a lot of number crunching, is to try to understand are we beginning to see correlates of response, and are those Will that be different for different cancers? And so, I mean, this is a longer term project, obviously, but I think the first step, especially as far as investors are concerned, is, you know, do you have positive results? And then after that, can you enrich for even more? Or is there some algorithm change that you can do to make the efficacy even better? And here, one of the things within Tissmaran that's really important is the enormous barrier that all that having all All that data and having it now. In front of other developers, for us, it creates an enormous opportunity to be the leaders and to maintain that leadership.

Akash Tiwari, Other

Understood. You made a passing comment at ASMO, but it was a very provocative one. You said, look, we are learning some stuff about renal. One of the things you did say was, we're actually doing a good job generating neoantigen, you know, not responses, but we are generating neoantigens. We just don't know, as you mentioned, with correlates, whether that actually correlates to efficacy. But can you expand on that a bit? What are you seeing with renal and, you know, the ability at least? Hit the target correctly.

Speaker 0

Yeah, you know, the thing with the big question with INT, just in general, is whether what we predict by algorithm in terms of mutation and how mutation translates to actual protein, which then is, you know, presented to the immune system through MHC, and that's HLA-dependent. So the question is, how do you know which one's the best one? And right now, there's a lot of trial and error. The thesis had been that, you know, those mutations that are, that can be, that are highly predictive to be expressed, that create epitopes that are completely novel, and that aren't going, that are well matched with the relevant HLA for the patient, And that's the ones that are going to do the trick, but that's That's a hypothesis. What was good about the renal data was that we were able to create Some of this neoantigens, because it's not TMB no, it's TMB low, and so we had that, but also to now test other kinds of mutations that might be relevant, but weren't necessary for the melanoma trial. So, you know, if it works, and I don't really don't have any data waiting at the data results when they come, we'll have a better, there'll be a richer repertoire of mutations that we'll be able to look at. And again, it gives you an opportunity to understand what's the aperture here and what's the work that needs to be done. And by the way, this is a great place where AI is helping us a lot because the amount of data that comes out of these studies is gargantuan, you know, over a thousand genomes and, you know, various immune responses and our choices and so on and so forth.

Akash Tiwari, Other

Makes sense. Last part on cancer vaccine, and I asked you this in ESMO. You know, you could argue the first cancer vaccine was Gardasil in many ways. Yes, it was. Uh, and there is this framework of, okay, there's a poster section question, you know, setting, but there's a limited patient group that would be there in a different cost structure. You are. Noting that you're exploring monotherapy and then combinations outside of Keytruda. What did you learn from some of the monotherapy data we've seen with the cancer vaccine about its ability to generate new antigen responses durably? And then what did you learn with that landmark analysis where, you know, it was really the, it does seem like the cancer vaccine is doing something, but it takes some time for that to really, you know, you know, like load up, I would say.

Speaker 0

Yeah, so, you know, I'll say that... You know, if we can create an INT that doesn't require pembrolizumab as background, it's going to be very helpful because, you know, Pembro is a great drug, God knows, and Merck knows for sure, MSD knows for sure, as, you know, MSD has been the leader here, but, you know, I have to say that when you start to look at people with very early stage tumors, you don't want to expose them to autoimmune phenomena and so on. So, you know, my preference would be to have, for the very early stage, have just the INT because, you know, the safety profile is relatively well-tolerated, just like injection site stuff. So I think we're, the proof, one of the biggest proofs, I think, will be in bladder and non-muscle invasive bladder cancer because there we have a pretty big experience that we're going to develop with INT only as well as with INT plus BCG. So that gives us some idea about whether or not you need that second immunostimulation. We also have a stage one, high-risk stage one case, but, you know, a phase three, but that's going to take forever to read out because it's still 10 years or so. I would say, by the way, the last thing is the prime boost stuff, the repeated priming, no surprise. That's, if you look at all... My, you know, my analogs is Shingrix and Gardasil, and, you know, you need to have continued priming before you actually get a magnitude of effect that's durable, right?

Akash Tiwari, Other

Actually, that will bridge to, I guess, a regulatory question, which is, you know, and this, we'll talk about this with PD-1 VEGFs, and, like, when the FDA considers accelerated approval or not, There's a lot of different facts. It doesn't seem like a bright line test. It's a holistic decision. And a big part of that is safety, right? So when I look at, let's say, you know, combining EGFR, TKI and chemotherapy, like, yeah, it'll work. But there's a toxicity burden. You could argue the same thing with a HIF2-alpha, Pembro, and renal. How does the discussion about accelerated approval change when you have a drug that has very limited additional side effects? To the background medication like Keytruda. You know, like, is there more receptiveness with the agency, especially for, let's say, elderly patients who can't tolerate some of these regimens, to get it on the market?

Speaker 0

Well, I mean, I think, first of all, you know, I don't speak for FDA, so, you know, or EMA or any of the other agencies. You know, I agree with you that benefit-risk is really important, both the benefit and the risk. And so that's critical. You know, when you have a background of pembrolizumab, you have a clear understanding of what to expect. There's, you know, the regulatory databases both in the U.S. and Europe and other countries is pretty rich. Adding INT to that really just increases injection site reactions. So I think that the calculus has to be magnitude of benefit, confidence intervals, and, you know, is this a clear signal, and can you take analogs from other areas and apply them? So, for example, Oftentimes, accelerated approval It can be facilitated by presence of data from other tumors, both from the safety database and the correlation between, let's say, an early signal versus a late signal. And so those are the kind of things you'd think about when thinking about AA. Unmet medical needs are also really important. And so, you know, it's a multifactorial thing, and you're right, it's a totality of data sort of thing. FDA has their systems. We never count on it. We just provide the data and see what these have to say.

Akash Tiwari, Other

Understood. Maybe bridging to the PD-1-VEGFs, and I was personally a little surprised that you guys made the announcement at Osmo about moving forward with, you know, like you talked about, we want to replace Keytruda. There was... Let's say an acknowledgement that this was a worthy backbone to combine with SAC TMD. And I think that has been, you've incrementally shifted your tone on this over a period of time. I find it really interesting that, you know, it was really now where, you know, you put your foot forward. How much of that was, you know, because you said Lenovo was a face-to-ready asset. So what did you see externally? That made you feel comfortable that it's worth adding a PD-1-VEGF to an ADC to really replace frontline standard of care and NSCLC.

Speaker 0

Yeah, so we're still in a, in the, we're doing it in a disciplined way and not in a Keytruda sort of try-everything approach, but, you know, we saw Harmony 6, which did show an OS benefit. We saw that there was positive results in biliary tract cancer, and, you know, Aqueso's data in China seemed pretty reasonable. Um, I'm still... We still need to make sure that this is for real, and I'm not talking about, you know, the early PFS or even early OS. I just want to make sure that cumulative toxicity isn't an issue, and it seems like it's not. I would also say that nothing helps more than having a lot of data on your own drug, and we saw we've had a pretty big program, again, in China where we do where things can be done pretty quickly in the early stages That led us to a very confident, to be much more confident about what we see with our drug, what's the dose that we would choose. And so then when we saw... The results of Harmony 6 and other studies like that, we see our own data, which looks really quite nice, and being able to achieve a dose that we think is going to be exciting in terms of ease of use. And then the SAC-TMT data from China, from Kehlun, made us realize that A plus B is going to be, you know, the two together are going to be great. We need to understand whether there's cumulative some sort of toxicity that needs testing, but that can be done in a parallel phase two program. We're also, I would also point out that we've had very productive discussions with FDA. And, you know, the way that I would just say that, you know, it's clear that There's been a lot of interest in moving this forward on their part, so I think we'll see. Proof is in the pudding.

Akash Tiwari, Other

That is true. That is the one thing about this. There's always a result. But we did get the Harmony CRL recently. And I think one of the things that stood out to me, look, we've seen the FDA-approved trials where maybe 70% of patients are in China. One of their sticking points seems to be not just patients, but number of events. So what did you – but you also have – you're looking at a lot of China data. You've relied on that. So what did you learn from that CRL, and how do you apply that to some of your discussions with the agency about your frontline strategy combining both of these assets?

Speaker 0

Right. So first of all, the U.S. is pretty clear that they want data that also includes U.S. patients, period. What they've been also clear about is that they don't love... And they said that in the CRL, they don't like studies that start off in China, enrolled most of the patients in China, and then you have kind of like a little add-on from international. So I think part of it, and so why is that? What makes that different than, let's say, a study that has 75% and 25% ex-China, but it's enrolled contemporaneously? The reason for that is the maturity of the U.S. data. The U.S. data are very immature. So they can't tell. Whether the overall data and the U.S. data are consistent, and you'll notice that a lot in clinical trials where the U.S. underperforms, and the reason for that is that it takes sometimes longer to enroll in the U.S. So what we've moved to is to U.S. front load the U.S. in our oncology trials so that you have mature data. It's not the percentage, it's the maturity that's important, and that really has made a difference for us, And I think we've had really good interactions with FDA regarding those elements.

Akash Tiwari, Other

Understood. I thought one of the, you didn't get as many questions on it as Mo, but I could tell your team was quite excited. The G12C data, I find that interesting on multiple fronts. A, it is an AI optimized drug, but B, That was a, I feel like it's a part of a broader take. You know, Merck's approach in first line with targeted agents are, I want a very clean, selective drug. I want to avoid chemotherapy. So it's not just about efficacy, but it's really about tolerability. You know, we've seen pan-KRAS agents, which are, you know, they have beautiful efficacy, but they do have a tolerability burden. So here's really what I'm trying to get at. Is your view of the world, I'd rather go in a frontline setting, targeted, selected agent, ideally chemo-free, versus, let's say, we take a pan KRAS agent frontline because it has a broader signal, right? Like, how are you thinking about the KRAS space eventually evolving over time for both lung and other indications?

Speaker 0

I really think that the credit should go where we should give credit where credit's due, and that's to AZ and to Grisso. So, you know, I mean, I have to say that, you know, a drug that you can take once a day that has a, you know, it's not a walk in the park, but it's a decent safety profile, and you can take it for years, and patients really do well, that to me is now there's a good drug. And so I don't see any reason why you need a pan KRAS or panRAS in lung because, you know, it's not difficult to get. You know, the mutation status, so I'd much rather have a tolerable drug, an oral drug, that can be administered with real inhibition and hitting hard the mutant KRAS. Calderasiv, or MK1084, is exactly that. It's really well tolerated. In combination with Pembro, we have really, really durable efficacy, I mean, really long-term efficacy. Why bother with chemo? You see the almiracid data for Lilly, and they actually had lower responses. So what we want from driver mutation cancers is to be able to give a drug that's actually going to allow patients to live a relatively normal life. And it's for that reason why we have a G12D that we licensed, and we have our G12C. And I think that the trick will be to have a really potent... It's a narrow-focused drug that will be well-tolerated and can be given as oral, oral with Pembro, which is, you know, pretty easy to do with Culex now, and so it completely transforms care for those patients, and that's what really matters.

Akash Tiwari, Other

Just to sneak it in, okay, so that's a framework for lung. Let's apply that to pancreatic, because that's where, you know, it's maybe harder to move away from chemo. Do you feel like that same approach is the right one in first-line pancreatic cancer?

Speaker 0

No, I mean, look, right now, pancreatic cancer is in a very different space. I mean, you're starting from, it's a death sentence, and you have drugs now that really make a difference. But if you look, you know, there was Ben Sasse, you know, had the senator from, the former senator from Nebraska, former president of University of Florida. He's been writing a blog about his experiences with pancreatic cancer. And, you know, he... He said that he felt like his face and body were melting there for a while when he was on the pan-ras inhibitor. But he's alive.

Akash Tiwari, Other

He is alive.

Speaker 0

And he was alive and continues to, had continued to do well. So that's the metric there. And, you know, ultimately, though, we can do better across the board. That's why we do research. We try to improve things and hopefully get to a point where pancreatic cancer is like what melanoma is. Melanoma used to be a death sentence. Now it's not at all. The same is true now. Hopefully we'll get there with pancreatic cancer and with lung cancer.

Akash Tiwari, Other

Understood. This is a wonderful discussion. Thanks so much for everyone for joining us. Elio, thank you so much.