All right. Good morning, everyone. I'm Stephen Lilley, one of the senior biotech analysts here at Stifel, and glad to have us as part of the next session, the CEO of CompuGen, Aran Afir. Aran, thanks for joining us today. Nice, Tim. Any opening comments or introductory statements you want to give before we jump into Q&A?

So maybe consult a few words, high-level description of CompuGen. So CompuGen is a pioneer in AI-based computational discovery. We were doing actually AI before all of this recent hype of AI. And we use this AI-based platform to identify new drug targets in the field of immunology. And we have a validated computational platform. And why it's validated? Because we showed again and again we can use computational tools to identify new drug targets. and we have internal assets that we'll describe together in the discussion today that were discovered using a computational platform. And the platform is validated by the collaboration we do with pharma companies around some of our discoveries and we'll discuss it as well with Gilead, AstraZeneca. We have, let's say, one critical milestone with the internal assets in the Q127 for a COM701, a block of PVIG, a unique checkpoint that we identified, very different biology, very interesting signal in ovarian cancer with the readout of a small randomized study in Q1-27. We have a collaboration, again, and we'll discuss it around the bispecific antibody with AstraZeneca, which is an 11-phase free trials. Another collaboration with Gilead, another asset, GI-0321, a cool approach to harness cytokine biology for the treatment of cancer. And behind all of that, we have a whole pipeline with innovative approaches to activate the immune system against cancer. And all of that is derived from this computational engine that we use to identify drug targets.

All right. Very good. So maybe we can start with the adaptive platform trial that's evaluating single agent COM701. Again, this is the anti-PB rig antibody. body. You're looking at this as maintenance therapy and kind of second line plus platinum sensitive ovarian cancer patients. Can you just remind us kind of what the framework of the trial looks like here with respect to patient numbers, randomization, endpoints, and then maybe just any kind of specific baseline characteristics that you're targeting for the purposes of patient enrollment here.

Sure. So maybe we can start by two sentences. What brought us to this study? So PVRIG, as I mentioned before, is a unique checkpoint with a very different biology, very different from PD-1, definitely from TIGIT. And from the beginning, ovarian cancer was an target indication for us because of the biology, of specific biology of PVRIG. And in general, the pathway is very dominantly expressed in ovarian cancer. And then we started the clinical trials, typically done, last-line patient, platinum-resistant ovarian cancer patients, and we saw and we presented in ESMO last year a pooled analysis of COM7-1 in monotherapy and combinations, and the signal was quite clear. We saw activity of COM7-1 in these last-line patients and in monotherapy and combinations. The patient who did get clinical benefit, it was very durable. The safety profile is excellent. And then we decided that for this type and then with discussion with KOLs, that for this type of drug with the safety and durability and mechanism of action, we should actually go earlier in the ovarian cancer treatment algorithm. And what we identified, together with the KOL, that in the second, third line, platinum-sensitive ovarian cancer patients. So these are patients which still respond to the platinum. They are much earlier than the retreated patients. Some of them were after 10 prior lines. So they are much earlier in the treatment algorithm. They have less exhausted immune system. We take only patients which are platinum-sensitive and responded to the last platinum-based chemotherapy. So they'll have low tumor burden, and we exclude patients with liver metastasis because we discovered that these patients are doing much better in response to COM701. So we're increasing the probability of these patients to respond to COM701 by selecting these specific earlier line patients. And there is a huge unmet need for these patients because in the second third line, platinum-sensitive of brain cancer, Patients who already received BEV or PARP are not eligible for BEV or PARP, there is no standard of care. So there is a huge unmet need. This patient received the platinum, they respond to it, but soon enough they will relapse and eventually become platinum resistant. So to find a drug that prolongs the response to the platinum is a huge unmet need, and this is the myovarian trial. We're taking second, third light patients, platinum sensitive, responding to chemotherapy without liver metastasis. We are randomizing them in a blinded randomized study. 40 patients will receive COM701 monotherapy. 20 will receive placebo. Nothing because, again, no silent of care, so we can do that. And the readout is going to be a progression-free survival with the readout in Q1A27. Yeah, this is the Maya trial.

And so you're excluding patients who have baseline liver mets. This is not uncommon for other IO-based studies that we've seen in other tumor types like colorectal. What proportion of patients would you expect to screen out on the basis of having liver mets at baseline?

So we don't have the exact numbers, but overall, our estimations, also with discussions we have, that in these relatively early-aligned patients, it's probably not much more than 10%. So it's not going to be huge patients excluded. But again, we think that it would allow us to further focus on patients that are more likely to respond to COM-701. In the PROC settings, in which also the higher frequency of patients with liver mets, We had like four times more clinical benefit in patients without liver metastasis compared those with liver metastasis. That's why we're doing this exclusion, even though we go earlier.

Maybe you can talk about your assumptions for progression-free survival in both the placebo and the COM701 arms and maybe specifically speak to your level of confidence around the placebo arm assumption and, you know, maybe how that assumption is sensitive to the type of clinical benefit that a patient achieves with platinum therapy.

So we're focusing only on the patient which responds to the platinum, so either complete or partial response. And based on the older trials, mostly the PARP trials that were done in these settings, in second, third line ovarian cancer, a patient to respond to the platinum and then receive maintenance. So the range is between 3.8 to 5.8 months after platinum, after finishing the platinum, before the patient relapse. So this is the median expected PFS.

So this is our expectation that will be somewhere in between this range, 3.8 to 5.8 months. and so i know that you've uh you're guiding to interim data here from this first analysis occurring in the first quarter of next year um will this be a analysis that determines either a no-go decision is is is there a futility assessment associated with this as well Absolutely.

Our goal here is to see a clear-cut monotherapy signal for COM701. There is a futility analysis. And in general, we want to see this decision point for us in general. If the signal is positive, not borderline, clearly positive, seeing monotherapy signal, then the first thing that probably we should do because of dynamite need is start discussing with the FDA about the registrational strategy. This will be the first thing. And then because this is an adaptive trial design, we can also now after proving COM7-1 is active in ovarian cancer in monotherapy, we can also build the whole strategy around, for example, we can test now one arm, add easily an arm combination with ADCs that are emerging agents in this field. We can add an arm with PARP and then consider going into first-line maintenance. We can also think, in general, going into last-line patients again, now with Pembro and Ocytaxel, which is recently approved. So after proving, and not to talk about other indications, which we have seen signal. So I think after clearly showing COM7-1 monotherapy signal, first we go for a gestational strategy in this patient population, but then the expansion could be quite broad in ovarian and outside of ovarian cancer.

And presumably the safety data you've seen would give you a lot of flexibility on kind of what you choose to partner with.

Absolutely. Yeah, I think it's both in terms of safety, also in terms of the mode of action. For example, ADCs, the mode of action is extremely complementary with the mechanism of action of COM701. Same goes for other potential agents. So, yes, combination. After proving the efficacy, the combinability of this agent should be quite strong.

And is the thought process around the potential synergy with an ADC, is that driven by this notion that, you know, these agents are highly cytotoxic? and as a result of their activity, you almost kind of get this MSI-high, TMB-high type of signature that then primes a patient's ability to respond to something like COM701?

I'm not sure it's exactly going to transform a cold tumor to become like MSI-high, but it definitely will induce some signals and antigens, and specifically the biology of PVRG, and we'll not go into all the details, but dendritic cell biology, antigen representation, stem-like memory cell interacting with disease, this specific biology can work very nicely with the signals generated by ADCs and cytotoxics in general, but yes.

Okay. Maybe the last COM701 question, and then we'll get to the earlier stage pipeline. How are you guys thinking about the addressable market opportunity here that exists in this in kind of this second-line-plus maintenance setting for COM701. I know that there's a decent amount of epidemiological data for second-line, but then I think as you get beyond second-line platinum-sensitive, that data is a little bit less clear. So what's kind of your best guess as to what the TAM looks like here from a patient numbers perspective?

Yeah, so for that specific population, of patients who received Bev or PARP or are not eligible for Bev and PARP. So these are typically second or third line platinum-sensitive patients. Our estimation is roughly around 8 to 12K yearly. But then again, by adding more ARMS, for example, if we just add Bev to this combination, this is another combination we can do, then we're already getting most of the second and third line. And then the numbers increase to, I think, around 30K. So the initial patient population, probably around 10K, but then we can easily go broader, as I mentioned before.

Okay. So you also have the GS0321 partnership with Gilead. This asset's currently in phase one dose escalation in solid tumors. Can you just speak to the progress that's been made on this program thus far? And, you know, how engaged are you with Gilead around the progression of this program and what next development steps might look like?

So really in short about the program itself. So I think this is another program that we identified computationally. It's not only in first-in-class assets. It's a first-in-class approach in which we harness cytokine biology for the treatment of cancer. but with a very differentiated approach using an antibody and not an engineered cytokine. And we think and show pre-clinically that this approach can overcome many of the challenges that others had and are having in these attempts to engineer cytokines to use them as the anti-cancer agents. So this is a really cool approach and a different one, and it's very exciting to take this kind of approach with the clinic. And it's even more exciting to do it with a great team like Gilead. There is very tight interaction. The teams are working together. We are leading the phase one trial, even though the asset was licensed to Gilead, but we're leading it. But it's tight interaction and a very good collaboration for this trial. In general, the way this trial is built, it's solid tumors, all comers, dose escalation in mono or in combination with Zim, the PD-1 of Gilead. Then in the dose escalation, we also have a few backfill cohorts in different doses. and then we have moving the phase the second part of the phase one study is those expansion in selected indications which is also run by us and by the terms of the collaboration after we end the phase one then Gilead will take it to phase two and further on and because of that collaboration obviously we cannot disclose specific details on where we are standing but we did start more than a year ago we dosed the first patient at the beginning of 25 and the trial is progressing really nicely. And hopefully with discussion with Gilead, we will report results a bit later when the trial progress a bit more.

Maybe to that last point, will you be able to present some of the dose escalation data before you move into dose escalation? I know sometimes these larger partners aren't really incentivized to give you bite-sized pieces of data that often benefit a smaller company like yourself?

So it's a good question. I mean, we have very good collaboration. Gilead are aware of the fact that they're working with the biotech that have this need to disclose data. And we are also are sensitive to their needs. Looking at other collaborations, they did, I think that they did allow in some cases to release also early data.

So we cannot commit for now, but definitely in discussion with them, I'm optimistic that we're able to release data sooner than later. but again we cannot commit okay and maybe you can just kind of talk to the longer term economic leverage um that you have to the success of this asset and um if there's anything that you can say about the potential cadence of milestone payments and whether there's meaningful milestones that are that could be earned within kind of these early stages of development so for this deal we received to date $90 million, received $60 million up front and $30 million after achieving the

IND clearance, were eligible for additional $758 million, and up to a low double-digit tiered royalties. About how the milestones are splitted, we cannot say much. What I can say is that typically, this kind of collaborations, when you move from stage to stage, there are milestones. And as I mentioned before, the phase one is split into the phase one A, those expansion, and phase one B, those escalation, and then those expansion, and then phase two. But you cannot give much more details on exactly where and how much we'll get the milestones.

Okay. The other partnership that you have, which is probably generating a lot more investor attention right now, just because it's been made a lot more visible, is Rilva Gostamig with Astra. This is a TIGIT PD-1 by specific. Astra now has this in 11 phase three studies. They've attached 5 billion plus of peak sales guidance to this. um i know tidget has been a pretty disappointing target which might be the understatement of of of the year uh we recently saw arkis and gilead discontinue the the dom program um you know what what gives you and astra you know the confidence that narilva can can can change the narrative around this target, TIGIT?

So this is a very important question. And now we're not speaking on behalf of Astra, so I will say what I think, being the one company discovered TIGIT and studying TIGIT for so long. So there are three aspects for that. It's the format of the antibody they use, the bispecific antibody, the combination strategy, and clinical strategy. And I will explain. All the failures, first of all, most of the failures of the TIGIT with the FC active TIGIT antibody, terrible safety, difficult to combine let's park this for a second because Arcus also failed with FC reduced the FC reduced gives you the option to combine very easily and still I think it should be more efficacious than FC actives, but then also comes the bispecific portion AstraZeneca have shown pre-clinically but in a very translational system of material taken directly from patients ex vivo system, they have shown that the bispecific, probably due to its cooperative binding, is different than just PD-1-tigit combination, and actually it's more efficacious than PD-1 and tigit combination tested thus far. So one, the bispecific by itself should be more efficacious, at least based on this data, and the clinical data will have to prove it. This is one. Two, which is also in part because of the bispecific, the combination strategy. When I have a bispecific that blocks PD-1 and TIGIT, first of all, the burden of proof for contribution of components could be easier in some of the trials. And again, going back to the FC-reduced, this allows you to combine, and you look at the strategy of AstraZeneca across the 11-Fa3 trials, they are really combining many ADC combinations with Enerto, Trop2 ADCs. And in some of the studies, they don't really, the way I see it, They don't really have to prove that TIGIT is active as part of the bispecific to get the win in that specific trial. For example, there was a phase-free trial of RILVE plus chemo versus chemo. They use it as an IO backbone, as a Keytruda-like molecule, if you may. And this trial is not testing TIGIT activity at all. It just relies on the fact it's a safe, in the worst case, very, very good PD-1. I think that it does target TIGIT, and it is more active than just PD-1. Other trials that are doing, for example, Rilve, Trompeo Lung 10, Rilve plus Trop2 ADC versus Pembro in the Trop2 selected biomarker patients. I think there's good probability of success regardless of a TGIT being active or not. And again, I think that we know that TGIT is active. The question is, is it active enough to gain FA3 success? So this is the combination strategy. And lastly, is the trial design. I think the way I see the way AstraZeneca designed their trials, that they learn from the mistakes of others, they are powering the studies in the right way to fit for the magnitude of effect that TGIT can give. They are focusing more, for example, ARCUS in their recent failure in the non-small cell study. At least they started as an all-comer study. They did adjust along the way to focus on PD-L1 positive. I'm not sure about the numbers. AstraZeneca are selecting PD-L1 to begin with. They're looking at squamous versus non-squamous separately. Each study nicely powered. So I think also the clinical strategy is learning for the mistakes of others and potentially fits more the biology and the magnitude of effect you can expect from TIGIT. So it was kind of a long answer, but I think it's covered nicely the three aspects, the format of the antibody, combination strategy, and clinical strategy, that along the 11th phase rituals should allow at least to some of them, if not most of them, to be successful as the way I see it.

Yeah, no, it's a good point, right? Because I guess where the TIGIT translation has failed has been in that, you know, phase two to phase three transition where you then need to show superiority versus PD1 alone. And that's kind of where a lot of these assets have fallen short. And the way Astra has designed these studies, you don't necessarily need to show superiority relative to Pembro. And you can lean on some of the other agents that they're including in the combinations to get you that win. I think we're going to be getting some updated data from Astra at the upcoming ASCO meeting. Anything that you think is worth highlighting to investors in terms of having important read through to what they're doing in phase three?

So there is one study, the ISPA, in combination with N-Herto, and this is in breast cancer, which they don't have in a FA3 study ongoing, but I think that there are some other interesting studies, FA3 studies of Rilve in combination with this amazing blockbuster drug, N-Herto. So I think still signs of efficacy versus historical data and the safety and combainability are going to be important and some read through to the N-Herto combination studies that Rilve is doing in FA3. and then there is hepatobiliary study not randomized study 30 patients i believe but still there is an ongoing equivalent phase three study so it will be interesting to see rilve plus chemotherapy and there we don't know if there will be os data or just pfs data but it will be interesting to see some comparison in terms of safety and efficacy compared to historical mainly the durva control in the top study which is the control arm in the phase three study

I know you recently monetized a portion of the royalties that are owed to you from Astra with Astra. Can you maybe just kind of talk about the rationale for that transaction? Why was that the right time to do it? And how does the structure of that transaction kind of, you know, preserve your longer term leverage to the success of this, of this drug?

Yeah, so this was, we did it at the end of 25, and the rationale was that we monetized a very small portion of the royalties we had. So we had mid-single-digit tiered royalties before the monetization, and we remained eligible for mid-single-digit royalties after. So we really monetized a small portion, meaning that if we will be the bulkbuster, we'll be living, and we still have full potential to receive these mid-single-digit royalties. and for this amount, we received $65 million up front and additional $25 million for the next milestones. So totally, we're eligible for additional $195 million. And while receiving this cash, now we have the cash runway into 29 and this will allow us to achieve internal and external milestones and to be there to receive potentially the next BLA acceptance milestone from AstraZeneca. So I think this was a great combination of giving us the runway to continue to invest in our pipeline, the early pipeline we didn't discuss, but we have the computational engine. We continue to invest in the early pipeline. So DishCache runway was exactly on time to allow us to continue to execute aggressively on all fronts and to be there eventually to enjoy potentially from a real-vigostomy clinical success in phase three.

Okay. Then maybe just last question, you know, the earlier stage discovery efforts, the computational platform, um how big of an effort do you have on going here in the background and um you know how do you think how do you think about the balance between capitalizing on bd opportunities like you pursued with gilead versus you know building out a wholly owned pipeline um that you guys can lay claim to so we invest huge efforts in the early pipeline we have this engine it's validated one And we believe that the assets coming from this agent can really make difference to patients.

So the largest team in Crumpogen is actually the one who works to continue bringing such assets. And about how we see keeping it internally without licensing, I think that now we have, when we have the financial stability, we have the freedom to decide. In the past, and as many typical to biotechs, you have an asset, you're running out of cash, you have to out-license it. The Gilead deal was an amazing one, $90 million, et cetera. But now we have the freedom to decide. we may want to keep the asset to ourselves if we think that we can generate more value to investors by keeping the asset maybe generate the phase one data or more but we also have the flexibility to decide okay for this asset if we think that the right thing for this asset is to go early on to pharma companies and to move it aggressively in the clinical pipeline we can make the decision so i think now with the financial stability we also have more flexibility to make decisions and potentially to maintain more value and keeping more assets and not necessarily licensing out early.

All right. That's all we have for time, Aron. Appreciate it, as always. I'm sure our paths will cross in Chicago here in the next week or two.

Absolutely. Thanks, everyone.