Oh, very good. Well, welcome back to Oppenheimer's 36th Annual Healthcare Life Sciences Conference. I'm Leland Gershell, one of the analysts on the biotech research team here, and really thrilled to have Compugen as our next company that we'll be featuring. Compugen is public, ticker CGEN. We've been covering them with an outperform rating for some time. We have President and CEO, Eran Ophir, with us, who will be joining us for a discussion on the company's programs and opportunities. If you have any questions you'd like us to include, please submit those through the portal. With that, just wanted to thank you, Eran, for joining us.
Thank you, Leland, for having us.
Good. You know, you have a few, you know, pipeline assets and programs. You know, you've got the partnership, obviously, with AstraZeneca, but, you know, internally, your focus really is in, you know, the maintenance ovarian setting. I think what we most want to hear about is kind of, you know, your current development there with COM701. You know, what's the rationale, biologically, clinically? You know, what do you think we might see from the randomized interim that you've guided to, I think, coming about a year out from now?
Absolutely. PVRIG is a target that we identified, which is a checkpoint with a very, very different biology compared to other, all other checkpoints, PD-1, TIGIT. That biology has shown to really enable us that upon PVRIG blockade, you can drive T cells also into less inflamed tumors. We have seen it research-wide, we've seen it in actual patient samples. Specifically, PVRIG pathway is very dominant in ovarian cancer. It was a target indication for us to begin with. This unique biology is one very important aspect of our COM701 asset, really a first-in-class approach. We identified computationally with our AI-based engine, this is really relevant for a tough and relatively less inflamed indication like ovarian cancer.
When we started our clinical studies, we were glad to see that this biology actually really translated into clinical signals. We had patients with ovarian cancer, PD-1 negative, typically not responding to checkpoints in all, that responded to PVRIG blockade, to COM701 monotherapy, and then also we've seen the signal in combination. In the last line, platinum-resistant ovarian cancer patient, this is where we always start, right? In the last line patients, we saw clear signals, monotherapy and signal in combination, that made us to believe, "Yes, this biology is actually translating to clinical signal." After doing a certain amount of operation, the substation, the signal in the last line patients, we had a decision to make.
We've seen that in the last line patients, many patients with rapidly progressing tumors to a IO-based only combination, like the one we used, the monotherapy, it was a bit, I would say, too harsh. Some of the patients were doing extremely well, safe therapy, durable responses, but some of them, the tumor was progressing before they even had the opportunity to respond. One opportunity was to add something cytotoxic and to stay in the last line patients. What we identified is the unmet need in the earlier line of treatment. What we identified that in the second, third line, platinum-sensitive ovarian cancer patients, these are patients which are still sensitive, much earlier than the one that we treated before. They're still sensitive to platinum. They received already bevacizumab as maintenance, or they're not eligible. They received the five to six cycles of platinum.
There's no standard of care, good biology, the clinical signals, the safety and durability we have seen in the last line patients in these earlier settings to prolong the progression-free survival of platinum-sensitive patients. This is a huge unmet need and a great opportunity, especially for a drug like COM701, with the durability and safety signals.
Great. What should be most important as we look for efficacy here? Is it the depth of response, durability, any subsets that might be defined by biomarkers, you know, given the PVRIG mechanism?
What we really look to see is, will the durability and activity we've seen in subset of patients, the last line patients. I will give you another piece of data we've seen in the last line patients, and we reported it in ESMO last year. We take the last line patients, the platinum-resistant patients, and we exclude, retrospectively, patient with liver metastasis, the clinical benefit rose dramatically. Already, only by doing this relatively simple exclusion of patients which are more, I would say, immune-compromised because of the liver metastasis and because that they have a higher tumor burden, probably, we just exclude those.
Already in the last line patients, we see almost 40% of the patients with clinical benefit, and the ones with clinical benefit, it was very durable, 10.5 months PFS, which is extremely long in these settings. What we are looking now in the platinum-sensitive in the Maia study, in this randomized study, in which we are taking 60 patients, which are after platinum, we randomized them into a blinded study. 40 patients will receive COM701 monotherapy and 20 will receive placebo. There is no standard of care, so this is ethically to do because they have no other option. The goal is really to see the durability. These are patients who already have low tumor burden when we start because of the response to the platinum, and we want to maintain the PFS.
We expect the control in this case, based on historical control, to be a PFS of 5.5 months, and we expect COM701 to prolong this ideally by three months. This will be really very clinically meaningful, and to really give these patients the option to prolong the response and not to relapse again and again and become platinum-resistant.
Yeah. In the past, you've seen, you know, durable responses with 701 and in some combo studies, I think patients who previously progressed on checkpoint inhibitors. Are the patients in MAIA-ovarian, have they been exposed to some extent on PD-1s, or what's the patient population like in addition to kind of, you know, platinum responders maintenance?
Yeah. Typically, these patients, because there's no standard of care, PD-1 approved in these lines.
Yeah.
there is now a recent approval for pembro plus chemotherapy in the last line, patient with platinum-resistant patients.
Right.
Typically, these patients are not going to be exposed to PD-1 prior to entering the trial.
Yeah.
They will be exposed to platinum, probably most of them to Bev, bevacizumab, and PARP inhibitors. Yeah, and we expect them to enter the trial in with low tumor burden, mostly responding to the chemotherapy.
Okay, good. I think you had shifted the readout a little bit from later this year to early next. Was there anything there, just enrollment, event rates, anything that factored into that?
It was not even enrollment or events. It was mostly about a problem which is not only ours, but it took a bit more time to open the big academic centers in the U.S. a bit more than anticipated. First of all, we're glad that now all the sites are open, especially the big sites, with all the support of the KOLs that actually are the one who drove this study. We're also glad that our Cangi Ginecò, which is a French group, which have expertise specifically in platinum-sensitive trials, they approached us to open trials in France as well. Now France is open as well. We have the sites open in France, U.S., and Israel. Everything is moving on track to meet the Q1 2027 guidelines interim analysis.
Got it. Okay. Good. Yeah, we look forward to seeing those data. Looks very interesting. Are there any other efforts in the industry that are going after this target, this PVRIG, that you're aware of actively?
There are a few early programs. There was one, PVRIG TIGIT bispecific, that BioNTech, it's an Fc active, so I'm not sure exactly the right format for PVRIG, but still they are pursuing that, and then some other earlier assets. We identified this target. We're the first to move in the clinic, and definitely the first to take it into a randomized trial.
Yeah. Okay, those are earlier. Yeah, so touching on TIGIT, so obviously, you know, investors are very aware of some disappointments in the past with other campaigns in the TIGIT space. You know, I think it would be just good for people maybe who are less familiar with your differentiation to hear about kind of, you know, what do you think the drivers have been for those who maybe have not succeeded versus, you know, rilvegostomig, which is, you know, you have partnered with AstraZeneca, which has, you know, the TIGIT component, but maybe differentiated, of course, from the other candidates that have been frustrating in their development.
Yeah. Obviously, this, there are big debates and a lot of frustration around TIGIT, around biotech investors, and obviously understood, given the failures that we all have seen. Most of the failures, I mean, for TIGIT, you have, at least for monoclonal TIGIT antibodies, and we'll talk about rilvegostomig in a second. For monoclonal antibodies, you have two main formats, the Fc active and the Fc-reduced. Almost all the failures have been of the Fc active. The Fc active has the worst safety profile, they're difficult to combine, they have the risk of depleting effect of T cells, and eventually, I think it's was not the right way to go to begin with.
We chose, like Arcus, like AstraZeneca, the Fc-reduced, which have, first of all, the safety for sure is very similar to the combination of PD-1 plus TIGIT. The safety looks quite similar to PD-1 alone, and this is very important, especially if you want to start combination trials. We think also this should drive better efficacy. It was obviously disappointing to see the first phase III readout from Arcus out of the three phase III that they are running, failing also the Fc-reduced, and we have to see how this evolves. For rilvegostomig, it's much more than just the Fc-reduced. Fc-reduced is a very good start to be able to combine and result with the right format. Rilvegostomig also is the only, until now at least, the bispecific antibody that was tested.
It's not a PD-1 TIGIT combination, but rather TIGIT PD-1 bispecific. There are a few advantages for a bispecific. One, the way it actually blocks TIGIT, it's a cooperative binding approach, blocks TIGIT and PD-1 on the same effector cells. Actually, AstraZeneca have shown that this approach, compared to PD-1 TIGIT combination, is more efficacious, at least in ex vivo patient sample, but in a very translational ex vivo system of non-small cell lung cancer. First of all, in terms of efficacy, there are biological and some proofs to think it will be more effective than the PD-1 and TIGIT combination that was done until now. When you have a bispecific, it gives you really the ease of combining it, sometimes without even having to prove the TIGIT combination, the contribution of components....
Looking at the way AstraZeneca, and I'm not talking on AstraZeneca's behalf, but I look on what they do and the way they combine, and they're really using rilvegostomig as a IO backbone that they combine with their pipeline. Some of the studies are doing direct head-to-head TIGIT PD-1 bispecific versus pembro, but other studies are actually using it as a backbone, then combine it with ADCs. I think the combination with the inerto is very interesting, and whether you're combining rilva plus inerto to pembro plus L16, you're giving your trial a more probability of success, not only because of TIGIT. Overall, they have 11 phase III studies ongoing, and I think that's really they learned probably from the failure of others.
To summarize, I think that the format of the bispecific, the type of combination that they are doing, and the clinical trial design, all of this gives AstraZeneca a good probability of success compared to the failures we've seen from TIGIT until now.
Yeah. Yeah, and AZ is always touches on rilvegostomig in their quarterly updates. They have a slide where they show, you know, all the various trials, atropine, and so forth. Clearly they're investing, and they continue to invest pretty heavily in this. I think they've even come out and said maybe a $5 billion +, you know, market opportunity, which is obviously very appealing. I'm wondering if, Eran, you want to offer your view, kind of what you think is maybe the highest probability path to initial registration success, you know, among the tumor types and lines of therapy, among those different development paths.
It's difficult because I think that's, you know, obviously, each of the phase III studies are being planned carefully, selecting the right patient population. For example, the ongoing non-small cell lung cancer trials are PD-1 selected. We know that Arcus is gonna have a readout in the first half of for their non-small cell trial, but it's PD-1 not selected. All of the studies are carefully planned, carefully powered, and a very interesting approach. What is unique to AstraZeneca, maybe that was also noticed until now, with the TIGIT, all the ADC combinations. Some very interesting study with the inerato, which is a very, you know, very promising drug by itself. The TROPION-Lung10 study is very interesting.
They are combining rilva with their Trop-2 ADC, also utilize their unique Trop-2 AI-based biomarker approach that enables them really to focus on the patient who will benefit the most, and to compare it to pembro. I think there's a really interesting approach to utilize the most of the Trop-2 ADC in combination with rilva. They just announced now a new study, it's actually in ClinicalTrials.gov, of a combination, phase III study in combination with Claudin 18.2 ADC. I think the ADC studies are very interesting, but also the others are carefully planned and powered.
Yeah. As you, as you think about, you know, different types of benefit and magnitude of the PFS or OS, or maybe just response durability, you know, what do you have in mind as kind of clearly differentiating for rilvegostomig from existing, you know, either PD-1 based doublets or some of the emerging, you know, ADC IO combinations? You know, is it, is it fair that between you and AZ, there are certain internal kind of go, no-go criteria, that they want to see before committing to further development or other large programs?
Yes. First of all, I mean, at this stage, AZ are taking this program by their own. We are not part of the decision making, and again, important for me to say, I'm not speaking here on their behalf, but they have our asset as part of the PD-1 T cell bispecific. I personally think that at this stage, with the 11 phase III trials, it's a bit late for a go, no-go.
Yeah.
I think it's a go. Eventually for them, and also obviously for the investors that are skeptical about TIGIT, eventually they will have to need to show in a phase III trial that rilvegostomig can score based on the primary readout of that specific study. It could be PFS, could be OS, could be both. Eventually, they will just have to score in a phase III study. Again, out of the 11 studies, I'm sure that some of them will be there. We'll all have to wait and see.
Great. If you could remind us, Eran, on your financial terms with AZ, you know, down the road.
Until December, last year, we received $30 million, and we're eligible for $175 million and mid-single-digit tiered royalties. What we did in December 2025, we actually monetized a small portion of our future royalties to AstraZeneca. We received additional $65 million upfront in December, and we also added $25 million for the next milestone. We are now eligible for additional $195 million, and the next milestone will be BLA acceptance, hopefully in 2028, even though they didn't give the exact timing, we cannot commit for that. Most importantly, after the deal, even though we monetized and received $65 + $25 near term, we retained the majority of the royalty interests. Even after the deal, we're still eligible for mid-single-digit royalties.
Overall, and you mentioned previously mentioning about rilvegostomig being a potential blockbuster of more than $5 billion annual sales, so mid-single digit out of that will be very meaningful for Compugen. It really will allow us to get a constant stream and financial stability and continue to utilize our AI-based computational engine to bring more assets and more collaborations and more clinical development.
Excellent. Yeah, let's spend some time on another very interesting asset that you've partnered with Gilead. This is GS-0321. It has the Gilead nomenclature. It's an antibody against the IL-18 binding protein. I think that you got into the first patients about a year ago in the phase I trial that's ongoing? If you'd be able to just provide us an update on the status of that, and you know, dose escalation, expansion potential, and when we might see some, you know, see a look on its clinical progress?
Yes, maybe two words about the asset itself. It's, again, a first-in-class approach that we identified computationally with our AI-based engine, Unigen. No one ever did something like that with IL-18BP. Nobody ever did this approach in general in oncology. What we identified, that in the tumor microenvironment, there are high levels of a cytokine, a very potent cytokine called IL-18, which is expressed there in high levels because induced when inflammation, very low in the blood. This is a key point of the discovery. It's there, not active because of an inhibitor for the IL-18 binding protein that binds IL-18 with high affinity and prevents its activation.
Overall, what we identified, and this is, we could do that because of all the database that we have with the tumor microenvironment from patients, that in the tumor we have this complex in high concentration of IL-18, IL-18BP, which is doing nothing. It's not active. We developed a high-affinity antibody that inhibit the inhibitor, displace the IL-18BP away from IL-18, and by that, enable IL-18 to drive its activity in the tumor, and because it's very low in the periphery, we're doing exactly the opposite from what other cytokines are doing. We're modulating the tumor and not the periphery, but where normally cytokines, which are great immune stimulants, but terrible drugs, are doing exactly the opposite. This is a bit about the mode of action, which is cool and new, and it's exciting to bring this into patients.
We started indeed, a year ago. Unfortunately, as per the terms of the agreement, as typical with pharma companies, we cannot give too much guidelines. We can say that it's moving exactly as we planned. We received until now $30 million for the IND clearance. We cannot disclose the exact next milestones. The trial is divided into two parts, dose escalation and dose expansion. There are some backfilling the dose escalation. We had a trial-in-progress poster in ASCO for the details. It's moving nicely, and hopefully, at a certain point, we will disclose results. We have to coordinate that with Gilead.
Yeah. You know, some people may have some concerns given the role of the IL-18 pathway in terms of, you know, immune-related talks or around systemic inflammation. Remind us, did you show pre-clinically that, you know, these are concerns that did not really appear in earlier work? You know...
Absolutely.
Between you and Gilead, you know, is there, you know, is there, are the monitoring requirements of that study such that, you know, they might be able to, you know, catch, let's say, any observations before, you know, they become an issue?
Absolutely. The role IL-18 play in inflammatory, not immune disease, actually is a great proof of concept for us, why the cytokine is a great cytokine to target. The patients on our trial obviously don't have any autoimmunity, so the sites in which there is inflammation is actually the site of the tumor. This is exactly the concept, that we are modulating the tumor because we have high IL-18 in the tumor, and we don't have IL-18 high in the periphery, because these patients typically don't suffer autoimmunity. It's typical for IO trials, we are excluding if there is concern for an additional site of inflammation, like a viral infection or any kind of autoimmunity disposition. Typically, as in any IO trial, we're excluding these patients.
The patient which arrive to the trial are the ones who should not have other sites of IL-18 upregulation. It mostly occur in the tumor, and therefore, what we have seen in our pre-clinical studies, that in this case, we're modulating the tumor and the blood, really nothing much happened then, especially in no toxicity. I think by doing the right trial design, we exploit the benefit of IL-18, and like in any trial, there are some risks, but I think that they are really mitigated nicely by the exclusion criteria.
Great. Okay. Yeah, we'll be patient for Gilead to provide us with updates there. Maybe if you could just again, remind us with Gilead, what the economics, how those are structured?
Yeah. For this asset, we received the... Still in the preclinical stage, because of this unique biology that overcome many challenges that other are facing with cytokine therapeutics. There was a competitive approach, and Gilead, eventually we licensed to Gilead, they received $60 million upfront, $30 million upon IND clearance, and here we're eligible for additional $758 million in milestones and up to low double-digit royalties.
Excellent. Okay, good. Yeah, you know, you've been successful with your discovery engine. You've got the computational technology, also a lot of expertise in the DNA access. Are there, you know, could there be announcements coming out of Compugen for, you know, new IND-enabling programs, target selections, anything along those lines we should be on the lookout for?
First of all, yes, you should be on the lookout. Unfortunately, we didn't give guidelines yet. For the early pipeline, usually we commit for guidelines when we are really when the asset is fully established and we are sure, and definitely it's when you're exploring new biology, it takes time. All this financing that we're getting from our collaborations are being definitely used extensively to continue and execute on the early pipeline. The biggest group with Compugen is the early pipeline. We're looking to new biologies. I would not necessarily wait for additional epigenome modulator. we have the expertise, we used it in the past, but can you already see from COM503, it is completely different asset, different biology. Our computational engine is built to give us the flexibility to look in different biologies.
What I can say is that we're working on different biologies. It's definitely not necessarily going to be anything similar to what we did until now. Moving forward, obviously, at certain points, we'll give guidelines, but we'll wait for more validation and more new biology established.
Excellent. Good, let me see if there are any other questions here. You know, one question is, some people have is, you know, some people were following Arcus. You know, they had a phase III trial with their candidate, domvanalimab, which was disappointing. Any commentary there from a Compugen perspective? Yeah.
First of all, it was disappointing also for us. we thought that the Arcus approach of Fc-reduced should be more effective. For sure, it's safer. The gastric trial indeed failed. We haven't seen data, so we're not sure exactly what happened, and we cannot say much because we haven't seen the data. We can say that AstraZeneca are not running any trial, which is in the same country, the same as the patient population, not phase III study, at least. They have now, in the, in the first half, they have another readout, or I would say, futility analysis of the non-small cell lung cancer trial, and we'll have to wait and see. If it's positive, obviously, it's great news also for COM902. If it's negative, again, it's a bit different from the AstraZeneca trials.
AstraZeneca trials are actually selecting patients based on PDL1, Arcus haven't done that. Probably, again, AstraZeneca have learned from the failures of others, they're designing those studies a bit differently. I think probably everyone assumes that it's going to fail, also, the non-small cell lung cancer trial. If it will be positive, it's a great upside. If not, I don't think it reflects much on the AZ trials because of all the reasons I said, because of the bispecific approach and which is different from what Arcus are doing.
Yep. Yeah, very good. Just lastly, I mean, many of you know, us biotech people go to these big conferences like ASCO, ESMO. Should we be expecting anything from Compugen? Maybe a little bit early for you to say, but just curious. In the past, you've, you know, had some presentations there.
Yeah. For now, we didn't give any specific guidelines. Again, for AstraZeneca, we know that there will be some readouts this year. Again, they didn't give us specific guidelines. For now, neither for us or for AstraZeneca, specific conference that we can point, it's a bit early.
Yeah. Okay, good. We'll be patient and look forward to for future updates. Thank you very much, Eran, for joining us. It's been a great discussion. Thank you all for Zooming in and hearing about Co
mpugen, and hope everybody enjoys the rest of the conference.
Thank you, Leland.