All right, welcome everybody. I have the pleasure of hosting Terns Pharmaceuticals here. My name is Zaki Molvi. I'm standing in for Amy Lee. We're all associates on Akash Tawari's team. So I have Amy Burroughs and Emil Kuriakose from Terns Pharmaceuticals. We're excited to host them. I'll let you guys give some opening remarks, and then we'll go into the Q&A.
Great. Thank you so much, Zaki, to you, to Akash's whole team, and to Jefferies for having us here, and for all of you who are in the room with us. Terns Pharmaceuticals is a company started in 2017. We're focused on small molecules with validated mechanisms of action. We're excited to have two clinical data readouts that came out in 2024 and two data readouts in 2025 for our CML program, our oral BCR-ABL inhibitor for chronic myeloid leukemia, where we believe that we have a best-in-class molecule that will be transformational for CML patients.
We have an oral GLP-1 receptor agonist for obesity that will also read out in the fourth quarter of this year, where we showed 28-day data last year that we believe this molecule can lead to competitive efficacy with a better tolerability profile and much simpler titration, which is important in this primary care market that is vast with many patients who are seeking the convenience of orals and a more convenient titration profile and something that's tolerable.
Great. All right, I think one of the questions investors are trying to understand is, I think you've said that Terns is a CML company first and GLP-1 second. Obviously, last year, the stock traded more so on the GLP-1. What has kind of led to the apparent prioritization shift?
Yeah, so it's really a capital allocation decision, Zaki. I think a lot of companies in this business evolve over time as clinical data emerges and as market data emerges. If you think about a small biotech like Terns, we have a really special asset in Terns- 701. This is an asset that has a very high probability of reaching patients, of being transformational. There is a very clear clinical and commercial and regulatory path to market. This is something that's well within our capabilities as a small biotech to execute on. In terms of capital allocation, in my mind, it does not get much better than that. In the obesity space, we've been in metabolics for seven years, which has gotten us to a place where we are in, I would argue, in the lead pack of oral GLP-1 receptor agonists with a potentially best-in-class molecule.
In the future, we see this as a partnerable asset. With differentiated data, we see the phase three, the commercialization, as more of a big pharma game. From a capital allocation perspective, we plan on really pursuing CML as a company and partnering out our metabolic assets that we have taken to proof of concept.
Got it. Makes sense. In terms of CML, we'll get an update in Q4 this year. I think so far, what we're seeing is a lot of people want to make the comparison to Enliven as kind of the most comparable competitor. Right now, so far, it looks like you've shown optically similar data and potentially better data than Scemblix. What can we expect to see at the six-month data cut, particularly that would help us understand what the differentiation is?
Yeah, so there was a lot in that question. Let me address a couple of pieces of that. Enliven and Terns are both enrolling a contemporaneous patient population in the phase one. However, we really view our benchmark as asciminib or an allosteric inhibitor, different from Enliven, which is an active site inhibitor. We really view Terns- 701 as a frontline drug and a second-line drug when an allosteric is chosen. We put out initial data in December of 2024, which was arguably the earliest data set anybody's ever seen in CML because we're a small biotech and we wanted to show that we've got a drug here. In fact, that data looked as good or better than we could have expected. We saw clinical responses in patients who had failed asciminib. This is, by the way, at our lowest two doses.
We have now said we've completed dose escalation in January. We also recently announced that we're moving to expansion at the highest dose of 500 milligrams because we did not see any dose-limiting toxicities and we saw a good safety profile across all four of our doses. We are also expanding at the 320 milligram dose, which is the next highest dose that is different enough from the 500 to meet the project optimist requirements. At the end of the year, what you're going to see in the fourth quarter is you are going to see a data set with a sufficient robustness to look at the six-month major molecular response rate, which, by the way, is the registrational endpoint in CML.
You're going to be able to compare that across other trials to look at the response rate for Terns- 701 versus other assets in development and those that have come before us.
Got it. Is it fair to say, I mean, Enliven has showed like 36% MMR in the post-asciminib failures? I think what people will be looking for is to see, is there signs that at your go-forward dose, you could do better than that? Number one, do you internally feel that is the bar, at least in this setting? Number two, is that actually the important data point if you're looking to go into earlier lines?
Yeah, so with regard to that question, I mean, we'll see the data from Enliven, I think, in full at EHA. It's important to distinguish cumulative MMR from MMR achieved. That 36% we think is referring to a cumulative MMR number. MMR achieved is the more important number in terms of determining the activity of the drug. That's also the regulatory endpoint. That's specifically referring to patients who at baseline do not have MMR, so a BCR-ABL transcript of 0.1% or higher. The MMR achievement rate is how many of those patients get to MMR or better by the six-month time point. Cumulative MMR is those patients plus the patients who came in at MMR at baseline who maintained that MMR through six months. We think the real bar is MMR achievement.
We know historically that asciminib achieved a 25% MMR at six months. That's MMR achievement that was in an allosteric naive population. We think that an MMR achievement of 25% in an allosteric and ponatinib-exposed population would be a very good signal of the drug being competitive with asciminib and other drugs. 30% achievement or better would be a great outcome, we think. That is sort of the way we're thinking about the benchmark. That is probably the more important benchmark in terms of MMR achievement versus cumulative MMR.
Got it. That's really helpful. I think at EHA, you guys showed, I think, in specific detail, your coverage over different mutations. I think one of the things that people might be missing is that the actual resistance mutations post-Scemblix are relatively low frequency. We're seeing they're like 3%-5%. I think you've kind of mentioned enrolling a population of just post-Scemblix failures who've acquired resistance would actually be really difficult. Can you talk about, one, the importance of actually the mutations that you do cover and what that means for moving up in lines?
Yeah, exactly. Great question. I think you hit it on the head. The target development population for this drug is not asciminib failure or specifically mutational resistance failures because so far from what we know from the asciminib long-term data, especially the frontline data, which is now approaching four years, the most common mutations that they've seen is the A337V mutation, which is a myristate pocket mutation, which is hovering around 4%. That's not by far the most common reason that people have to switch. The real unmet need is there are patients who still can't tolerate asciminib. Despite it being a great drug, there are still issues with that with regard to things like amylase, lipase, hypertension. These patients are on these drugs for life. These things do matter in terms of even low-grade events.
Seeing a differentiation there would be very meaningful with 701 because in the real world, more patients do switch with allosterics with the Scemblix intolerance than true resistance. We think that we can overcome that. Resistance is not only defined by progression. It is also defined as suboptimal response. As we showed in our December data, at our lowest dose, there was a patient who had not gotten a response to ponatinib, then went on to Scemblix and flatlined their response after having a one-log reduction, went on to Terns- 701 at the 160 mg lowest dose, and deepened their response within four cycles to another log. That patient remains on therapy. Those are the kinds of responses that we want to see in that post-Scemblix population.
Just zooming out, the majority of these patients are going to have seen a mix of all the two GTKIs, Gleevec, and several seeing asciminib and ponatinib. I think in that population, again, achieving that MMR rate of 25%-30% or higher would be very meaningful in terms of how this drug performs on an efficacy standpoint.
Got it. One of the important things we've kind of seen is that there does seem to be a dose response for Scemblix. If you were able to dose higher, you could push the efficacy farther. It looks like at your highest dose so far, you're not seeing DLTs. What kind of data can we expect to kind of flesh out that thesis further that you could differentiate on efficacy just by having, as you're saying, a more tolerant drug?
Absolutely. That is the premise of the allosteric class as a whole, that because they are so selective for the fusion protein, because of the uniqueness of the mirror state pocket, you can get a much wider therapeutic index compared to any active site drug. So far, fortunately, that is sort of what we are seeing with the data. We have not seen any DLTs all the way up to the max dose of 500, where we are seeing several-fold higher coverage over the target than the in- vivo exposures where we saw inhibition of BCR-ABL and downstream signaling. That would argue, based on historical precedent, that when you have a patient with imatinib, for example, at 400, and then they sort of flatline the response, they kind of do deepen when you go to 800. Higher doses do matter, but those drugs are limited by safety.
Seeing that good safety profile through the dose range is a really encouraging sign for us that we could really leverage that wide therapeutic index and dose these patients at the higher doses where you can maximally deepen responses.
Right. One of the other things you've kind of mentioned is this is an open-label trial. You do get to hear some investigator feedback sort of in real time. I know Emil and me both were at Sloan Kettering probably at different times. I was in the basement. He was probably seeing patients. What kind of feedback are you hearing from the trial centers in terms of the profile? It does seem like the trial centers overlap with your competitor as well.
Yeah. I mean, we've been very pleased to see that our enrollment in this study has really taken off over the last several months, which is why we have very good conviction that we'll see a very meaningful data set in the fourth quarter. Investigators, when they see a drug that looks safe and it looks like it's working in patients who have gotten pretty much all the other options, they put more patients on study. That's sort of the feedback that we're getting across all the sites. We're running a global study with sites in the U.S. and Europe and Australia, South Korea. To your point, there's a good amount of overlap with the other phase one studies in CML that are ongoing as well. We're seeing really good enrollment. We're very encouraged by that.
Got it. In terms of your ability to move rapidly into first line, when can we kind of expect? We will get the six-month data update at the end of Q4. When can we expect to hear an update on your ability to move in first line? How will that kind of timelines in terms of starting enrollment and expectations on first line data sets?
Yeah. Look, we have a very sort of real-time back-and-forth discussions with the FDA and other health agencies. For example, because of Project Optimus, the doses that we selected for expansion were reviewed by the agency. Given that active dialogue, we think that based on the robustness of the data in Q4, and if we have a good sense of sort of where the numbers are with regard to MMR and knowing the safety data as well, that we could zoom in on sort of the dose that we would think as a go-forward dose, we would start to have conversations with regulators as we'd expect to think about what a pivotal could look like. To your point around frontline, we've said that the end goal with this drug is to have as wide a label as asciminib currently has. That includes frontline and subsequent.
Obviously, you have to get there in steps. We think that the first step to get there would be a second-line plus trial based on borrowing data from this phase one, which, by the way, already enrolls second-line patients. That would be sort of a seamless next transition to what would be a randomized study in a second-line plus population. There is only one step to get to a frontline population with the study that probably looks not that different from ASFA first.
Okay. And so you would be going head-to-head versus the two GTKIs or?
Correct.
Okay. Got it. Okay. It sounds like there would be an intermediate second-line plus trial, and then you would do first line. In terms of when looking at data like that or profile that we can expect, we think the differentiation for Terns is you have the tolerability angle that allows you to get to higher exposures. You have the PK, which looks like there is not a food effect like there is with Scemblix. Also, there is a potential for less DDI. In your kind of what you are hearing in your market research in terms of what is important for getting switches, let's say, in the second-line setting or even moving into first line, how would you rank those in terms of what is most important?
In all of our research, and we spent a lot of time with both our KOLs and with community physicians, it's really a very complex decision process in terms of the different patients. For example, historically, before allosterics, despite second generation coming along and having better efficacy than imatinib, there's still been a lot of imatinib use due to the side effect profile. There are different comorbidities with each patient. All of these CML drugs have been successful with some different profiles, but not real differentiation in the second line. When you talk to physicians about how they think about it, for some patients, the food effect really matters. For some patients who say are younger, they really want to hit the disease hard and hit it with the most effective drug that they see.
We think that in totality, the profile that we're talking about for Terns- 701 is going to have pretty broad appeal when thinking about tolerability, when thinking about food effect, when thinking about efficacy when an allosteric is chosen. We do not see necessarily switching patients who are on a drug where they're tolerating it well and in MMR, but there are about 17,000 patients a year newly diagnosed with CML in the G7. An allosteric, we believe, will be the lead choice due to its efficacy and safety profile. With the qualities of Terns- 701, we believe that will be the best choice for patients who, for cost reasons or other reasons, may choose a first or second-generation agent. About half of those will end up switching in the second line. In that case, our physicians tell us we would switch classes to an allosteric.
For the same reasons, the Terns- 701 would be chosen in the front line, would be chosen in the second line. We see a real opportunity to gain significant share with a superior profile on different aspects.
Right. In terms of that kind of that first line, those 17,000 patients per year, how do you see the treatment shaping out when you have two GTKIs going generic?
Yeah. So we can actually look at analogs with when imatinib went generic and they were branded. We saw kind of fairly even share. We think that allosterics will have at least 50% share and potentially up to 70% or 75% share as the market evolves, as people gain more comfort with using allosterics. Right now, Novartis has said they have very good commercial payer coverage. There's been brisk adoption since the label was achieved in November.
Got it. And just in terms of, I guess, kind of so everyone can understand, in terms of an ATP site inhibitor versus an allosteric mechanism like yours, why should the therapeutic window be better for an allosteric?
It's basically because the mirror state pocket is not shared with other proteins and kinases. What's unique about it is usually that pocket's on wild-type ABL1. Wild-type ABL1 has an N-terminus that binds that pocket. When the fusion protein happens, that N-terminus is chopped off. The pocket stays wide open. Because that pocket is an autoregulatory pocket, binding that open pocket allows to shut the protein off without inhibiting other kinases that have a lot of homology with the active site. The off-target inhibition profile is that much more selective.
Got it. This isn't just pie in the sky because we've already seen you guys don't have DLTs so far at your highest tested dose. Yeah, personally, on our end, we really like the allosteric under way. We think there's a lot of potential there. I guess in this portion, I kind of want to talk about the oral GLP-1. There's been a lot of interest. We're going to get an update there as well expected this year. I guess talk to us about, I think the framing is you guys were using a danuglipron scaffold. I think the question in everyone's head is post-danuglipron's discontinuation, how should we be thinking about Terns- 601?
Exactly. I think we think it's a good thing for Terns- 601. It's one less player in the mix. We've shown that Terns- 601 has a really differentiated profile on the clinical side with our 28-day data. We showed very linear dose response with regard to weight loss, achieving the high end at 5.5%. We saw a very good tolerability profile with very rapid titration, arguably the fastest titration of any 28-day study with oral GLP-1s with zero discontinuations, dose interruptions, or reductions.
We think that's due to the molecules' unique properties that allow for a very nice flat PK curve, prolonged absorption that leads to 24-hour target coverage, and a combination of high gut exposure relative to plasma, giving substantial activation of the target in the gut, which is, I think, what matters from a site of action standpoint, and a very high protein binding and low plasma free fraction that gives us that tolerability profile. The other key aspect that people often overlook with these is we have very simple titration, right? I mean, I think doing all the research we have done in this space, specifically also with primary care physicians, where we think is going to be the bulk of this market, they want a drug that achieves reasonably good weight loss.
I don't think anybody in the primary care setting, especially in the mild to moderate BMI category, wants to lose 20% of their body weight in a year. These patients are fine with a weight loss that approximates mid-10% to teens weight loss over a year period. They prioritize tolerability above all else and simplicity. What I mean by simplicity is it would be very easy if you had a GLP-1 that you didn't need to titrate, period. That's the ideal scenario. If you had a GLP-1 that only took two or three steps to get to the target dose and that was super tolerable, that's much more, I think, much more attractive to a primary care setting both for clinicians and patients. That's where we think this molecule is going to differentiate. To your point around danuglipron, I think we're now one of the handful.
I could count on one hand the number of companies in the small biotechs with clinical stage oral GLP-1s that are currently in phase two. We are very excited about the phase two 12-week data readout that we expect in the fourth quarter.
Right. I think kind of the read for danu is there was when they moved to that kind of that QD formulation, they did not have any kind of issues before. Suddenly, when they started looking at the titration optimization with their new formulation, they started to see some signs that made them concerned that there was a High-low case. Why is not that the same as Terns- 601, which gets this short four- to six-hour short actual half-life, but then the apparent half-life is much longer and you actually get daily coverage? How can we decouple the comparison there?
No. Yeah. First, let's talk about danu, right? I mean, there's very little data on the details of those cases. We can only speculate here because we haven't seen the data yet. From a pure numbers standpoint, that drug's been over 1,400 patients. Numerically, we've seen these two cases. Obviously, they're real cases. We don't know if it's therefore accurate to just therefore flexibly attribute that to the scaffold. We don't think that's the right conclusion. Our speculation is that because of what you pointed out with the mix of the MR and the IR that led to a very different PK profile, namely, they rapidly titrated and then probably what we think has led to accumulation that increased exposure substantially above what you see with danuglipron IR BID. That may be what led to the liver tox. Again, we don't know.
Why is Terns 601 different even with that? Again, the key aspect of this molecule is that, like I said, the very high protein binding, which is meaningfully higher than what we know danuglipron's is, the low free fraction in plasma. Even at our doses of 250, 500 that numerically are higher than the other dose of danu, when you look at exposure to exposure, the exposures are lower than danuglipron. Add in the very high protein binding, the free drug exposure in plasma is a fraction of what you see with danuglipron. We saw essentially a flat liver profile with even up to the max dose with AST, ALT, total bilirubin in our phase one study with a dose range that's arguably wider than any other drug that's been evaluated in this space.
We don't see any concerns based on both the preclinical safety and the clinical safety data we have so far that liver tox is going to be an issue with this molecule.
Right. I think just to kind of make sure expectations are set properly, because we cover Lilly with orfo and type 2, we're hearing about fluctuations in AST, ALT that might be kind of normal just for patients who might be diabetic or prediabetic. In terms of thinking about that, in terms of, okay, whether you come in on baseline where you're already like 3x upper limit of normal, how would you actually interpret those kinds of cases?
I think that's valid that just by virtue of losing that rapid weight, you can see fluctuations in AST, ALT. In the study that the inclusion criteria for most of these studies, when you're above 30, you are not allowed by the FDA or the regulators to enroll patients with comorbidities so that you get a clean signal on things like liver. Our population is obviously not a comorbid patient population. That data will evolve as these molecules move forward. Again, based on what we know about this molecule, its properties, preclinical toxicology data, and obviously the clinical safety data. I would argue that there are other molecules that are also based on the danuglipron scaffold out there that we've seen clinical data for, and we haven't seen a really true liver signal across the board.
Again, the data will tell, but right now we do not have any concerns around the liver tox 601.
Got it. In terms of you mentioned you have pretty high absolute doses. You were going up to, I think, 740 milligrams in the phase one. These are high absolute milligrams, but because of kind of the high plasma protein binding and then the gut solubility, it kind of creates for a smoother PK profile where you're not rapidly hitting Cmax and getting a really high peak to trough ratio like we might have seen with danuglipron, which kind of complicated the tolerability profile. In terms of, I guess, on a longer-term basis and thinking about how the drug might accumulate, let's say in a longer trial, let's say phase three where you're going over 72 weeks, how are you thinking about the potential for AE incidents?
I feel like it's usually front-loaded, but then if you're changing the accumulation or changing the PK, that might be a different story.
We see no accumulation with the molecule. I'll point out that we are presenting some more of that, the phase one data at the upcoming ADA conference. We'll have a full PK curve, but essentially there's no accumulation, namely because of the fact that the PK curve and that 24-hour coverage is primarily driven by the prolongation of the absorption phase. Once the absorption phase is complete, the elimination phase is very rapid. Effectively, that results in no accumulation as you even go beyond several weeks or months of dosing. Obviously, we'll see the evidence for that in the 12-week data. I'll point out to your point about the dose itself, absolute dose, do not conflate dose with exposure, right?
Because like I said, even at these numerically higher doses, when you look at plasma exposure side by side with other drugs, we're on the lower end, yet still seeing very meaningful weight loss and good tolerability. That is sort of where this thesis comes from.
Got it. And how many tablets? I think you mentioned you do not have a G 4 dose selected yet, but might be around the 500-ish range. How many tablets is that going to be and how should we think about that in terms of variability?
We're thinking basically making 250 mg tablets. And at 500, that would basically be two tablets once a day.
Got it. Yeah. We'll be at ADA digitizing your PK curves to see.
I'm sure you will.
Yep. Okay. In terms of kind of running phase two, I think one of the things about phase one we're seeing with GLP-1s is that there's studies being run in inpatient settings, which you can kind of control for discontinuations a lot better. What are you thinking about in terms of phase two design and being able to show durability on weight loss? What kind of profile are you looking for ultimately in terms of balancing the efficacy and tolerability?
Our phase two design, the 12-week design, uses a slower titration, obviously, than we did in 28 days, substantially because our earliest up titration happens at two weeks into the first dose. Like I said before, the number of titration steps is the lowest of any 12-week design that we've seen, and we think we can lower that even further as we get more data. In terms of the types of data we're looking for, again, we prioritizing tolerability is what we want to really emphasize. We think that because of the slower titration, the good tolerability we saw already with very fast titration, we should see a very competitive tolerability profile with the best drugs out there combined with simplicity of titration.
With regard to weight loss, we think that we'll be in the range that every other drug has shown at 12 weeks, which is roughly in the 6-8% adjusted weight loss.
All right. We are just about out of.