Guggenheim, and I'm very pleased to welcome our next presenting company, Adagene, and with us is Mickael Chane-Du, Chief Strategy Officer. Mickael , thanks for joining us.
Thanks for having me.
So maybe just to start, for those who are newer perhaps to the Adagene story, could you provide a quick overview of the company, your platform, and your lead asset, muzastotug, also known as ADG126?
Sure. Adagene is a company focused on protein engineering, and we are notably known to develop a we have our own proprietary fully owned masking technology, and as you mentioned, we have a lead candidate called Muza, or ADG126. I'm going to say 126 for the remainder of the chat. 126 is a mask and a masking CTLA-4. We believe that it is a very differentiated CTLA-4 and a differentiated masking antibody. We will talk more about the data, but we believe that we have expanded the therapeutic window of CTLA-4 inhibition with it.
Great. As you mentioned, CTLA-4, proven IO target, but immune tox has always limited it to some degree.
What was built into Muza to widen that therapeutic window, and how is it fundamentally different from first-gen antibodies like ipilimumab?
Sure. Number one, we bind to a different epitope compared to ipilimumab, tremelimumab, or the botensilimab of the world. That differential epitope binding leads to a much stronger ADCC compared to Ipi. We believe that it is 10x higher, and this is data that has been replicated by third party, so we feel very good about mentioning that number. That's number one. Number two, we are using our SAFEbody, our masking technology to it. Two things about the SAFEbody technology applied to 126. A, we are protease sensitive, and we believe that our masking efficiency is best in class, which is very important. You don't want the CTLA-4 inhibitor to bind to the target when the mask is supposed to be there, and when the mask is supposed to be cleaved, you want it to bind to the target.
The whole concept about the masking efficiency is very important to us, and again, we believe that we have best-in-class masking efficiency. B, I'll mention also that we not only need certain proteases at high concentration in the tumor microenvironment to have activation and cleavage of the mask, we also need high expression of the target here, in this case, CTLA-4. So we need these two conditions for activation, which we believe participates strongly in the expansion of the therapeutic index.
Okay. So, you reported some very meaningful phase I update this past spring for Muza plus pembro in late-line colorectal cancer patients that do not have liver metastases.
Could you just walk us through the data you've shared to date and how that emerging clinical profile compares with the standard of care?
Sure. So I'm going to start with the safety. We have generated data for ADG126 in combination with pembrolizumab in the setting you just mentioned. Let's start with the safety. Our discontinuation rate due to adverse event is less than 10%. It's very close to what you would expect from KEYTRUDA single agent, even though we are dosing patients all the way up to 20 mg per kilogram. For people who are not aware, the first generation of CTLA-4 antibodies usually are given between 1 mg and 3 mg per kg. For example, when you look at ipilimumab, the fact that we're giving such a high dose without patients dropping out is very significant in our opinion. Number two, let's turn to the efficacy side, ORR, PFS, and OS.
At the relevant doses, which goes from 10 mg per kilogram Q3W all the way up to two cohorts at high dose 20 MPK, you have 20 MPK Q6W, and we have something called the loading dose, 20 followed by 10 with a schedule of Q3W. At these relevant doses, our response rate in the setting you mentioned is between 15% and 36%, and we're talking about confirmed responses. This is quite an important. This is quite a substantial range, keeping in mind that single agent KEYTRUDA is known to have a response rate close to 0%, and durvalumab, tremelimumab, in the very same setting, had a response rate of 2%-3%. That's in terms of response rate.
Our median PFS is four to five months, and importantly, at the combined 10 MPK cohort, we have generated a median overall survival of close to 20 months, followed by a plateau, which you would expect from immuno-oncology agents. We believe that that kind of dataset compares very favorably versus not only KEYTRUDA single agent, the first generation CTLA-4 PD-1, but also compared to the standard of care in the same setting. Again, setting of patients without liver metastasis.
Okay. Based on the data to some degree, you're also evaluating Muza now in a phase II study, which I think updated today has completed enrollment. Can you remind us of that study design and your line of sight into the first clinical update from the phase II?
Sure. This is a phase II that we're doing to comply with Project Optimus of the FDA. We have 60 patients randomized one-to-one between two doses. Dose number one is 10 mg per kilogram every three weeks in combination with KEYTRUDA. Cohort number 2 is 20 MPK Q6W, also in combination with KEYTRUDA. Just to be clear there, we're talking about dosing schedules that are on label for KEYTRUDA. The primary endpoint of that trial is response rate, and there's a number of important secondary endpoints, safety, et cetera, duration of response, that we will be looking at. In terms of timing, you should expect data in the first half of 2027.
Again, keeping in mind that the primary endpoint of the trial is response rate, but we obviously want to have a view also on what could be duration of response, early PFS, but it will be too early to make a call on survival.
Okay. At what point would you be comfortable pulling the trigger on a phase III? Do you need to see more maturity on the PFS or some early survival data, or is that early read sufficient given what you've seen already in phase I?
I think the early read should be sufficient. We plan to have a readout again in H1 2027. The next step would be discussion with the FDA at a phase II meeting to discuss the design of a potential phase III in the setting of late-line MSS CRC without liver metastasis. The official guidance is that we hope, we expect to have that phase III initiated in the later half of 2027.
Okay. How do you think about the phase III design? Would this be conceivably randomized versus LONSURF + Bev or a slightly later line with fruquintinib?
Mm-hmm. In terms of design for the phase III trial, you should expect a randomized trial, obviously, with overall survival as primary endpoint for the trial. There's potentially an upside scenario, and this is something that needs to be discussed with the FDA. We have noticed some precedents of the FDA allowing people to have overall response rate as a co-primary endpoint, and/or as a result could be used as a basis for an interim analysis and potentially an accelerated approval. But you should really think of it as an upside scenario. Okay. In terms of control arm, to be determined, but we have a pretty good view on what it could be. Again, when I say to be determined, post-interaction with the FDA. We believe that the nature of the control arm would be dictated by the line of therapy.
As you mentioned, LONSURF + Bev would be something that would be more relevant for a pure third-line population. When you talk about the fourth-line population purely or above, we are talking about the fruquintinib or regorafenib of the world. If you have a mix of the two population, third and fourth line, you should probably expect a physician choice of therapy depending on the line of therapy. Again, LONSURF + Bev for third line, fruquintinib or regorafenib for fourth line and above.
Okay. Then you would be theoretically seeking a label that restricts by patients without liver mets, which would be a first in colorectal cancer to date. How confident are you that you have the right alignment with the agency and physicians on carving out the segments as a potentially unique indication?
We feel confident about that based on our prior interactions with the FDA. We had a very solid end of phase I meeting last year, which led to the initiation of a randomized phase II for Project Optimus, et cetera. We have noticed also some good precedents about it. I mean, before they decided to pivot, Agenus did have a phase III called BATTMAN, looking at botensilimab, balstilimab, CTLA-4, PD-1, versus, I believe, physician choice of therapy for later line of CRC. The population was actually, no, sorry. The population did not include only patients without liver metastasis, but they also had phase II. They also had a phase II in there. Again, based on our own interaction with the FDA, we feel good about the label.
Okay. Colorectal is obviously a very rapidly evolving space potentially. Currently, standard of care, very poor, but there is a lot of new mechanisms, including ADCs that are moving into the space, but these are more for an all-comer population. How would sort of the Muza/pembro opportunity fit in with this evolving landscape?
Yeah. We believe that CTLA-4/PD-1 can offer a great option for the patients. Immuno-oncology agents are normally not known to induce a lot of responses, and they really win thanks to an overall survival benefit and what we call the tail of the curve, of the OS curve. We believe that with Adagene's 126 plus pembro, we can not only increase the response rate, but we can also have very durable responses, and all which ultimately can lead to very differentiated type of survival benefit and curve.
Okay. Makes a lot of sense. One thing that we've been really trying to pay attention to is the fact that the opportunity for Muza extends beyond just the late-line non-liver mets colorectal population, and you have a lot of other things ongoing, including a collaboration with Incyte in combination with their PD-1 TGF-beta bispecific. Yeah, would love to get your thoughts on the main opportunity for this combo and your view on the data that they've reported more recently at ESMO in an earlier setting and how that sort of changed your conviction on that.
Yeah. We're very excited about this trial collaboration with Incyte. For those who don't know, we signed this collab earlier this year with Incyte, and Incyte is sponsoring a trial to evaluate the bispecifics, again, PD-1 TGF-beta in combination with ADG126. The rationale behind it is very, very strong. PD-1/TGF-beta leads to a progression of CTLA-4 and Treg, and we are a CTLA-4 binder and a Treg depleter. Again, PD-1 single agent is not known to be super active. Their bispecific has shown what they believe and we believe what is a very differentiated kind of response rate, 23% in patients without liver metastasis and 12% in patients with liver metastasis. By adding ADG126, both companies hope that we will see improvement in outcomes, improvement in response rate, and so on and so forth. This is something that is allowed by ADG126 and the SAFEbody platform.
We believe that it would be difficult for a non-masked CTLA-4 to allow the right therapeutic index with their bispecific. Very excited about it. They have started enrolling patients recently. They've shown a lot of love, and we believe that the enrollment should be quite efficient.
And obviously, would they be the ones, I guess, driving the timelines for possible clinical updates?
They will be, yes. It is their trial. They are sponsoring the trial, so they control the timing of the data readout.
Yeah. Okay. Also in an even earlier setting, you are also evaluating Muza plus pembro in the neoadjuvant setting of colorectal, which is, as you mentioned, one of the areas that your peers has recently moved into for their PD-1, CTLA-4 combo. So, I guess first, what is the unmet need here in neoadjuvant colorectal, and how is a PD-1, CTLA-4 regimen positioned to sort of improve that need?
Yeah. Number one, there is technically no FDA-approved therapy for neoadjuvant CRC. If you look at the trial design of the ROBBIN trial from Agenus, you are looking at botensilimab, balstilimab given before surgery, and after surgery, you have watch and wait or chemo. The control arm is placebo surgery and then chemo or wait and watch. There is a very significant unmet medical need there. I mean, it does not mean that people do not receive anything before surgery. In the U.S., you have off-label usage of chemo, right? Based on the data from bot/bal, it seems that CTLA-4 plus PD-1 can induce a very high rate of responses and major pathological responses, which could be very relevant in terms of EFS down the road.
The data from Agenus seems to suggest that, but data from BMS for neoadjuvant also suggests that any kind of response, it could help the event-free survival of those patients, again, after surgery, et cetera, et cetera. We believe that in that setting, having a chemo-free regimen could be very, very appealing to the patients, given the differentiated kind of safety profile, on top of the fact that you can induce a lot of very deep responses.
Got it. Yeah. Agenus has been very optimistic about their opportunity there, and they are de-risking the path, if you will, in terms of regulatory front.
Right now, Muza is in the investigator-sponsored trial.
How should we think about when we could see an update from that study, to the extent that you can comment on that? Then, if you were to pursue this as a path, what would the next steps in theory be?
Yeah. In terms of timing, we have said that our IST in Singapore, so investigator-sponsored trial, should have data in 2027. We have said also publicly that we have amended the protocol for that trial, and we are now testing a dosing schedule that is very relevant, we believe, for that setting, which is something very close into the loading dose in our phase I-B, which again, is 20 mg per kilogram followed by 10 mg per kilogram with a Q3W schedule. We believe that dosing higher and dosing more potentially could lead to even better outcomes in that setting. Keep in mind that in the ROBBIN trial, Agenus bot/bal is given at 1 mg per kilogram in combination with their PD-1, and the CTLA-4 component is given for only one cycle. We would be giving more than one cycle, very high dose. We have a very potent CTLA-4.
But we need to generate the data, again, probably in 2027.
Okay. And where would the most relevant angle of differentiation be? Would it be driving deeper responses and sort of longer event-free survival, or is there a safety angle as well?
Safety is very important because the only potential cure for the patients is surgery. You don't want to derail that. You don't want to further delay it, et cetera, et cetera. But at the end of the day, it's really about efficacy. In oncology, it's all about efficacy, right? Having more responses, deeper responses on paper would be great. This is the bar that we set for ourselves in the neoadjuvant setting, but let's see what kind of data we can generate. Ultimately, obviously, event-free survival. But something that I would like to mention is that if you look at, I believe the name of the trial is NICHE from Bristol Myers Squibb for neoadjuvant, even partial responses, so we're not talking about major pathological responses or complete responses, but even PRs seem to lead to a benefit in terms of EFS.
We need to look at things holistically. You want to look at responsive, depth of response, and then ultimately, event-free survival, because this is the primary endpoint of a pivotal trial.
Okay. Could you potentially initiate a phase II study, for instance, after the IST data reads out, if it's supportive?
This is a possibility, yes.
Okay. Okay, great. Looking beyond colorectal, you have also generated data for Muza in other indications, including hepatocellular carcinoma this past year at AACR. Can you remind us of what you found in that study, and is this a path that you may pursue either on your own or with a partner?
We like first-line HCC a lot. It is an indication and setting that we consider very strongly. As a reminder, at AACR this year, we generated data, or at least Roche generated data in the randomized setting. We had an arm where ADG126 was tested at 6 mg per kilogram Q6W on top of the key standard of care, which is atezolizumab, bevacizumab, so PD-L1 VEGF. We showed in that small randomized phase I-B/II that we have potentially improved substantially the response rate, the PFS, as well as overall survival. To give a few numbers, on top of my head, the response rate for the control arm was close to 20%, and we had, in a small number though of patients, six, a response rate of 50%.
The PFS was close to doubled, and we had the trend towards better survival benefit versus, again, the control arm, with a very good safety profile. Atezo + Bev is, again, a key standard of care widely used for patients in first line. There is no approved regimen, including PD-1, CTLA-4, VEGF. We believe that Muza or ADG126 could be very safely added on top of that standard of care and potentially further improve the outcomes. Keep in mind that Nivo + Ipi, forgot the name of the trial, but generated a response rate of circa 36%, I believe, without Bev. Durva Treme, I believe, with one dose of CTLA-4, had a response rate of circa 20%. Atezo + Bev is known to be in the 20%-ish . We believe, again, that adding CTLA-4 on top of PD-1 VEGF makes a ton of sense.
Okay. I believe the data was generated in a sort of basket study or umbrella study, rather, for Roche. What would you have to see, I guess, or generate in order to have some movement in perhaps advancing into another study?
In first-line HCC, right? For first-line HCC, given the benchmarks that I just mentioned, Nivo + Ipi is 36%. The Atezo + Bev is in the 20%-30% range. Assuming that it is a very large sample size, having something that is in the 40%, 50% or above could be great in terms of response rate, but we need also to look at the PFS, OS trends, et cetera.
Okay. Makes sense. Beyond Muza, you have also been active in partnering around your SAFEbody platform in general. Can you remind us of some of the recent activity and perhaps any upcoming milestones that we could look out for?
Yeah. We have been very active on the BD front over the last few months. Remember, last year, we expanded our collaboration with Sanofi, and Sanofi made an equity investment. They are currently our second-largest equity shareholder. The relationship with Sanofi started in 2022 with an early-stage collaboration, discovery collaboration that we have since expanded. We signed other deals, for example, with Third Arc Bio. The Third Arc Bio shows you that we can apply the SAFEbody platform not only to monoclonal antibodies but also to T-cell engagers. We announced a first deal in 2025, and we a little more recently announced an expansion of that collaboration with them. We have been, again, very active on the BD front. We want to remain very active on the BD front.
We obviously cannot give any timing or nature of the discussion that we have, but we want to do more of what we have done in the past.
Okay. So then perhaps with our last minutes, could you quickly summarize your catalyst flow for the next 12 months, let's say, and your cash runway?
Our cash runway goes to the second half of 2028, so keeping in mind that we raised $70 million earlier this year after releasing updated clinical data as well as the trial collaboration with Incyte. In terms of catalysts for next year, expect Project Optimus phase II data that will help our interactions with the FDA and the future phase III in late-line MSS CRC without liver metastasis. So we said H1 2027. 2027, you should also expect data from our phase II in Singapore from this IST trial that we've talked about in the neoadjuvant setting of CRC. Something also to be determined or confirmed, but more recently, we announced that we had clearance from the Chinese FDA for an earlier stage compound called ADG138. It's a double-masked T-cell engager going after CD3 and HER2.
We will be enrolling patients in China first before potentially expanding the scope to something more global. But hopefully, we may be able to have some early clinical data for that asset as well next year.
Okay, fantastic. So I think with that, we're going to wrap up.