Eikon Therapeutics, Inc. (EIKN) September 2026 Conference Transcript

Good afternoon everyone. I'm Sean Armand, Head of US Mid-Cap Biotech Equity Research here at Morgan Stanley and welcome to the Morgan Stanley Global Healthcare Conference. Before we commence, to make you aware of some important disclosures, please go to the Morgan Stanley Research Disclosure website at www.morganstanley.com forward slash research disclosures and if you have any questions, please reach out to your Morgan Stanley sales representative. for this session we have the pleasure of hosting from ICON Therapeutics their CEO Roger Perlmutter and CFO Freddie Bowie thank you gentlemen for attending and look forward to the discussion thank you Sean very much and thanks for doing this I've got a couple of macro questions to begin with so let's just punch through them and get to the meat which is ICON so how is the rise of China originated innovation changing if at all your competitive positioning in your R&D?

Well, I think, you know, that the rise of China has affected everybody in the industry, of course. I see that as a natural evolution. This has been going on for a very long time. You know, 25 years ago, we watched as a flood of very high quality Chinese graduate students made their way to the United States and were studying in American laboratories or even more, even further ago than that. And it's only natural that many of those people, well-trained, experienced, decided there were opportunities in China that were superior. And there was a chance to harness an enormous amount of capability there. It's important to point out that the Chinese government has been investing heavily in this sector. And so it's understandable that you would expect to see a lot of innovation coming from China, and there is a substrate there that enables that to be done extremely well. I think what we've seen over the last 10 years is a progressive increase in sophistication in China, so that whereas before one had concerns about GXP execution, how well compliance worked in the clinical setting intellectual property issues but I think more and more we're just looking at another group that's very capable of producing high quality material to be studied for the improvement of treatment of human disease which as far as I'm concerned can only be a good thing we expect to be in a global business forever because around the world patients are experiencing the same kinds of problems and the more that we can do to improve treatment for such patients, the better off we're going to be. So I embrace it. I think it just makes us better.

Sure, wonderful. Thank you. And a little bit unusual asking you, John, when this question pertains to AI, and obviously AI is obviously embedded in your platform. But, you know, the question is, you know, beyond that, are you implementing AI across other parts of your business? and can you point to a definitive outcome on influencing a decision, a cost outcome, or a POS?

Well, you know, the reality is that AI affects every aspect of every business. Most of the work is quite prosaic. The fact that you have a trusted partner, if you will, that can write quite effectively, grammatically correct English that is more or less submission ready is of course saves an enormous amount of human time which is great most of that as I say is relatively prosaic work for our business as you point we little icon is one of the largest data producers on the planet and the analysis of that data simply can't be done by human beings. There's just too much data. When our instruments are running at full tilt, which they often are, we're producing on the order of a petabyte of new data a day, which, as Russ Berman, our lead engineer, says is a ridiculous amount of data. And processing that amount of data is only possible with the availability of extremely high-quality AI-enabled data analytics tools. We happen to have a lot of sophistication in that area. And that's what helps us with decision making. The actual decision making with respect to the more straightforward parts of the business, intellectual property, the way in which we manage finance and cost control, et cetera. Sure, AI assists on the margins, but it's not central to our decision making.

Sure. Thank you. And final question before we get into the detail around icon, and maybe the answer is all of these, but which policy variable are you focus on the most? Is it FDA, Medicare negotiation, tariffs, global pricing? What do you think about the most, Roger?

I think we're much more concerned. I'll speak for the company, but I'm really speaking in a way for myself. Much more concerned about the evolving commercial environment. um the the regulatory environment of course is always important but frankly most of the routine work in the regulatory environment is being done by people who have done it for a long time uh and there are there are more complex issues in regulatory science that sometimes it may be more difficult to answer uh but a larger question is how is the commercial environment really going to evolve with most favored nation policies, with the IRA, and with the tariff question, which affects, for example, even the conduct of clinical trials in a pretty dramatic way because there are tit-for-tat reciprocal tariff impositions that have an effect on your ability to bring new molecules into clinical trials. So those things concern me more, frankly, than the regulatory environment.

Thank you, Roger. To move on, so ICON 1001 in lung and the Telluride 005 study and thinking about the ESMO setup is my next line of questioning. So let's start with 1001 in lung and at ASCO you showed a 63% ORR and 91% disease control across the efficacy of valuable population with a non-squamous cohort out to a median duration of beyond 11 months at 13.6 months of follower. What do you see is the most important takeaway from that data set and how you think about the durability signal as it matures?

Let me start by just introducing to the audience what ICON's presentation environment will be like at ESMO because we have seven abstracts that were accepted at ESMO, which is a large number for any company and certainly for a small company like ours. We had six at ASCO. Every single one of our clinical candidates has clinical data that are being presented at ESMO, so it's a lot. The most advanced program is the 1001 program, which is the toll-like receptor dual agonist program, which is in registration-enabling studies in both melanoma as well as non-small cell lung cancer. For 1001, we will present mature data on the non-small cell lung cancer study, which you alluded to, the Telluride 005 study. And that mature data includes an additional five months, roughly five months, because the last data cut was May 4, an additional five months of experience. We now are past the two-year mark for many patients in the non-squamous, non-small cell lung cancer population. And in the squamous, this is the first time we've had the opportunity to present mature data involving all of the enrolled patients. Overall, I mean, of course, I haven't even seen the data cut that's going to be presented at ESMO. But I think we can feel quite confident that, once again, the data are extremely supportive of the idea that this drug candidate can be used in combination with Keytruda and chemotherapy, appropriate for histologic subtype, in a manner that's quite safe, extremely well-tolerated, and that appears to be providing efficacy both in terms of response rate as well as duration at the high end of expectation. And certainly that's gratifying, and I think there's a lot of enthusiasm in the thoracic oncology community for that. Wonderful. Thank you, Roger.

Given the competitive backdrop for PD-1 plus VEGF and TLR approaches in squamous lung, How do you think about 1.001's tolerability profile and combination opportunity, and how central is it to the eventual positioning?

Well, I think that the interesting thing about 1.001 is that we were so concerned about safety and tolerability. This is a potent TLR7-8 dual agonist. And, of course, we were very concerned about cytokine release syndrome. we were able to effectively manage that and at this point we've treated more than 600 patients and cytokine release syndrome is really not an issue we do see some grade 1-2 symptoms which could be secondary to 1-0-1 in the sense of fever and chills but of course you do get fever and chills from chemotherapy or from Keytruda alone beyond that we see very little in the way of symptomatology there's no reason why as a result we can't use 1-0-0-1 in combination with other platforms if it were to turn out that a VEGF dual antagonist which incorporates a PD-1 or PD-L1 warhead as well as a VEGF receptor warhead And there's no reason why we couldn't use 1, 0, or 1 in combination with that. We haven't done any such experiments. And at the moment, that landscape is so complicated, we could spend hours talking about the various different data sets, those that are China-only, those that are from the West, which combinations, et cetera, et cetera. I think we'll have to wait until the data play out.

Wonderful. Thank you. Overall, the response rate and duration are encouraging. How are you thinking about the path from this Phase II signal to a registrational readout in non-small-cell lung, and what role does Telluride 008 play in that?

So Telluride 008 is our registration-enabling study in non-small-cell lung cancer. It is a straightforward study design which simply compares the standard of care, which is Keytruda plus chemotherapy, for both squamous and non-squamous to that same standard of care plus 1,0,0,1 at the appropriate dosage schedule. Because of Project Optimus, we will have two doses, and we do have two doses in the run-in phase. And the first interim analysis collapses the dose. We did that in our melanoma study, Telluride 006, collapsed the dose just last month, and in a relatively short period of time, we expect we'll be at the same point with 008. At that point, it becomes a two-arm study and it continues on until such time as we learn what the results are. Our expectation is, of course, these are powered for PFS and overall survival in the appropriate patient populations. It will take some time to get those data, but certainly what we've seen thus far is encouraging.

Sure, thank you. And moving over to melanoma, so can you walk us through what the interim data told you and why you're comfortable with the dose and blinded data and what the second interim, the futility analysis, gating the phase three portion, is designed to establish?

Right. So, again, the trial design I've already mentioned. In this case, Keytruda alone is standard of care around the world for the treatment of advanced malignant melanoma. Our study was designed to compare that standard of care, Keytruda, versus Keytruda plus 1001. Of course, for Project Optimus, we have two doses at the beginning, and those two doses collapsed to one in August. We compared 0.6 milligrams per meter squared versus 0.75 milligrams per meter squared, and our independent data monitoring committee recommended that we proceed with 0.6, which is as we had hoped. I would hasten to add that these are blinded studies, and we don't see any of the data. The studies are under the supervision of an independent data safety monitoring committee, and so they make recommendations to us, and they made the recommendation that we collapse the dose to 0.6. The enrollment for the melanoma study has really been quite impressive. We had many more patients than we needed in order to make that interim call, but it takes time to assemble all the screens, et cetera, et cetera. So we're really far along in the enrollment profile. The next step is to ask the question of whether it is appropriate to proceed and enroll the entire study. And that is based on the judgment of the independent data safety monitoring committee they need to make that judgment based on event rates, as the whole study is really powered on event rates. So while I know how many people have enrolled, and broadly speaking, I can look at blinded data and see event, I really don't know what the event rates are, and they're going to have to make that judgment. And at some point, there will be sufficient events that we feel like, okay, it's appropriate at this point to do that second interim. the choice really for that data monitoring committee in its most formal sense is continue as planned or make some adjustment but within that those different categories data monitoring committees have a fairly broad brief they can their remit is pretty large they can choose to recommend a variety of things and we as sponsor will respond to their recommendation. My guess is that the recommendation will be to continue the study to completion because, of course, that's why we do these studies because we want to find out the answer. And, of course, the Data Monitoring Committee wants that to happen too.

Sure. And I guess you framed study completion around the first half of 2028.

Well, you know, the melanoma study, again, all of these are event-driven, And the melanoma study is powered on ultimately overall survival, of course. We do not expect that response rate will be a reasonable basis for registration in melanoma. So we would not imagine that in 2028 we would have sufficient overall survival data. That will take probably what we've said is in the first half of 2029. And that is because, of course, with Keytruda therapy, and certainly we believe Keytruda plus our 1001 molecule, we expect people to have fairly durable responses, and a progression-free survival should be reasonably long. And those event rates are going to determine how long it takes, so I can't answer the question in a formal way. But my expectation is that what we'd originally said is, gee, sometime in the first half of 2029 is probably about right.

Sure. Thank you, Roger. Moving on to the Part 1 selective franchise, and you've got some presentations coming up at ASCO. So what are you hoping to demonstrate, and why do you believe Part 1 selectivity translates into a cleaner combination?

Yeah, so the hypothesis is, as you say, pretty darn straightforward, which is that we had had the privilege of working with our colleagues. We, and I say that, Roy Baines, our chief medical officer, and I had worked together with our colleagues at both Merck and AstraZeneca to develop Linparza, which is the leading PARP inhibitor used for maintenance of chemotherapy-induced responses in patients with ovarian breast, prostate cancer, and some patients with pancreatic cancer. The problem with Linparza and all of the other dual PARP inhibitors is they have significant hematologic toxicity. And there was good evidence some time ago from preclinical studies that a selective PARP1 inhibitor would have less hematologic toxicity. We chose to explore that in a clinical setting. To do so, we scoured the earth, basically, to find the most selective PARP1 inhibitors we could. And we have special tools that enable us to do that that are part of our robust platform, instrumentation platform that we've built. Unsurprisingly, and to your very first question, we found those compounds in China, the medicine cabinet for the world, and a small company called Impact Therapeutics, which is run by a terrific organic chemist who I knew from a prior life, Dr. Tsai, had some great-looking molecules, we thought. We evaluated them, and we advanced those programs in both a brain penetrant and a non-brain penetrant fashion. The non-brain penetrant is the most advanced. It's called 1003. It's about 650-fold selective for PARP1 over PARP2 at the enzymatic level. The brain penetrant, even more selective, 800-fold. These are, to our knowledge, the most selective PARP1 inhibitors that are in clinical trials. They're also the most advanced selective PARP1 inhibitors in clinical trials that are truly selective. and encouragingly the safety profile for these molecules really does look different and we've presented this data in a variety of different settings previously at ESMO at ASCO and we'll have more data coming up at ESMO the data with respect to relative tolerability from the standpoint of myelosuppression was so strong that we did a study in which we combined full-dose peclitaxel at 80 milligrams weekly with 1003, given as a daily oral dose, for the treatment in the second line of ovarian or breast cancer in patients who had failed first-line therapy. That study sounds simple the way I said it, but the reality is it's never been possible to do a study like that before. We tried for years to do that with Olaparib by dose reduction for Parza Olaparib. But we could never get to a satisfactory dose and schedule. But we were able to combine 1003 with peclitaxel at full dose and full dose intensity for 1003 in the treatment of breast cancer where patients had failed either hormone deprivation or chemotherapy and in ovarian cancer where they had failed platinum therapy. And the results were really quite impressive. We presented these data at ASCO. We'll have the opportunity to update this still further at ESMO. And in addition to that, at ESMO, we'll be presenting data for the combination study with abiraterone, novel hormonal agent for the treatment of prostate cancer. And while I have not seen the final data, I would just say we're really encouraged by this molecule. We love this molecule, and it's behaving extremely well. So that suggests that we're in a very different place, and I want to emphasize this for the audience is that we're not just talking about the safety, improved safety profile of a selective PARP1 inhibitor. We're talking about the ability to use it in a very different environment. Rather than for maintenance of a treatment effect, as is typically done with all of the existing dual PARP inhibitors, we can actually go into treatment and treat patients in the second line, and we believe ultimately in the first line. So these are exciting data, important data, and I think there's going to be a lot of enthusiasm for these data at ESMO.

And with the brain penetrant, the 1004, Pub1 selective, how do you position that against 1003 within the same portfolio?

Well, there's a little bit of a time lag between them. 1003 is considerably advanced over 1004. 1004 is a more selective molecule. It is brain penetrant. we will be presenting data on 1004 for the very first time at ESMO. So no one has ever seen data on 1004, and that's an important thing. This molecule is extremely well-behaved, and I think it is reinforcing to see high-quality data on myeloid suppression in that setting, in 1004, as compared to 1003. It's worth remembering that in 1004, because it is brain penetrant, we could include patients who had active brain metastases. And ultimately, 1004 is also appropriate for potentially looking at the treatment of patients with primary central nervous system malignancies. So there are areas where 1004 can go, where 1003 simply cannot. On the other hand, there are opportunities where potentially we could track one or the other for this broad set of potential indications, which, again, cover breast, ovarian, prostate, pancreatic, et cetera.

So all of those areas are potentially available. Sure. Thank you. Still on ESMO, because there is a lot coming up from ICON, Sir, internally generated Wernickela case inhibitor, 1005. What do you hope to present? What should we expect?

Yeah, so 1005 is the first molecule that came from our platform that has been brought into the clinical setting. We've published data on 1005 before in terms of how this work was done using single molecule tracking as a means of optimizing the product candidate. And then we've also, because 1005 behaved extremely well and preclinical toxicology studies, we were able to take the drug into healthy volunteers in order to look at target occupancy and distribution, absorption, and, of course, the issue of duration of action. What we learned was it's extremely potent, which we knew, but the half-life is quite long. It's an orally bioavailable drug, but it has a 9.4-day half-life, which enables us to do once-weekly dosing. One of the things that has bedeviled the field is although it's clear that Warner Helicase inhibitors have activity, as has been shown from the original data, let's call them first-generation molecules from Novartis and Roche, the gastrointestinal toxicity was significant and problematic as well as some other pharmaceutical issues. we do not see that kind of adverse effect profile. In fact, our drug looks extremely clean. Armed with the information from healthy volunteers, we've gone into patients with both MSI high and MSS tumors. MSI high tumors are the ones that we expect to be most effectively treated. We will have a chance to present some of that clinical data. And our clinical program is quite novel in the sense that it is both as monotherapy and also in combination with a Keytruda checkpoint inhibitor, of course. So I think that looks very exciting, and we're eager to present the data. There's a lot for people to look at.

Sure. So the trial design is a little bit different. It sure is. And can you explain the trial design for investors and the rationale behind it?

Yeah, I think that our view is, look, checkpoint inhibitors are quite effective in the treatment of MSI-high cancers. It's among the most responsive tumors. But on the other hand, it's not uniformly that tumors are not uniformly responsive. And it's not clear that the right way to treat patients with MSI-high tumors is serially. In addition, there is both the MSI-HI that results from DNA damage repair defects, and there's also MSI-HI that is acquired as a result of other treatment protocols. So, what we sought to do was first establish the safety of the molecule by looking at 1.005 as monotherapy in patients who had failed prior checkpoint inhibitors. But in addition, we felt once we had demonstrated that, we could actually take a window of opportunity approach and look at MSI high patients who could be MSI high, but as a primary treatment, a run-in treatment in patients who would then have MSI status ascertained, we could then ask the question of whether in those patients, of course, they might see a response and they could be maintained on the therapy or thereafter treated with a checkpoint inhibitor. And in another arm of the study, we also could look at checkpoint inhibitor in combination with MSI high. And all of those kinds of approaches are beginning now in the cancer setting. It's an important way to give us an opportunity to think about how you sequence an active drug that works by a different mechanism in combination with another active drug that is already registered in the particular indication. And it gives us a chance to get a lot of that data assembled early on in the study. So we're quite enthusiastic about this approach.

Awesome. Thank you. On 1006 and the platform, so that's your androgen receptor program, it's moved faster than planned. You know, what made it move faster than planned and what drove the acceleration? How are you positioning it against established AR agents?

Yeah, so you're right. 106 had moved faster than planned, and our plan was very fast. I think one of the things that it's important to recognize, and I know we're getting short on time, but if I could, I would just tell you that the single most important source of value in our company is our discovery research platform. because it enables us to identify targets that elude others and to prosecute targets that others can't easily pursue. And we've demonstrated both of those things, and we continue to do so. In the case of 1006, we actually began the study because we were looking for novel inhibitors of the androgen receptor that would interact directly with the unstructured N-terminal domain of the receptor. We're still doing that, but 1006 came out as a ligand-binding domain inhibitor, but it had very different properties from the existing molecules like enzalutamide and darolutamide and others. When we explored that, we came to understand that, first of all, the molecule we identified is a non-lutamide. It binds differently. It's extremely potent, and we were able to, using our medicinal chemistry approaches, and this is all done using single-molecule tracking and advanced imaging, we were able to increase that potency dramatically, which helps us address one of the important mechanisms of resistance to existing androgen receptor blockers, which is amplification of the androgen receptor locus. And with greater potency, of course, you can manage that. Meanwhile, we could also optimize the compound to be effective in those point mutations that are selected for by drugs like enzalutamide and dialutamide, which is responsible for about 25% of resistance that occurs in treated prostate cancer patients. And in addition, we were able to edit out the partial agonism that exists in existing lutamide androgen receptor antagonists, which we could do using a machine learning model that we developed that could distinguish between agonism and antagonism using imaging as a method. That led to 1006. 1006, maybe unsurprisingly, in light of what I've told you, kind of sailed through its preclinical talks. It looks extremely good. We'd originally intended to pull all that data together and file it in 2027, but actually it looked so good that we said we can accelerate this and file, and we will file in the next few weeks, we expect, and then move very rapidly into the clinic. And there's a clear path here in treating patients with prostate cancer. And I have to tell you, a lot of enthusiasm in the prostate cancer community for this drug. So it looks really quite good.

Awesome, awesome. Maybe a question for you, Freddie. So maybe the balance sheet, and where does the cash balance get you to?

Yeah, thanks for the question. So at 6.30, we posted results and had $531.2 million of cash, not to put too fine a point on it, but a strong balance sheet. We're burning roughly $65 to $70 million every quarter. So you can think about that in terms of just runway moving forward. I will say Roger did a great job talking about our clinical work, But over two-thirds of our investment each quarter are in the clinical programs, and I do anticipate that to continue growing as those programs head towards continued registration and enabling studies and head, hopefully, to the market.

Well, we're just about out of time, so with that, I'd like to thank you for your time today, gentlemen, and thanks for participating in our conference.

Sean, thank you so much for giving us this opportunity.

Yeah, welcome. Thank you, everyone.