Compugen at Needham Conference: Strategic Immuno-Oncology Insights

On Monday, April 7, 2025, Compugen Ltd (NASDAQ: CGEN) presented at the 24th Annual Needham Virtual Healthcare Conference, highlighting its strategic initiatives in immuno-oncology. The company, which uses its UniGen platform to discover novel drug targets, emphasized both its promising pipeline and the challenges faced in advancing its therapeutic candidates.

Key Takeaways

• Compugen is advancing COM701 in a phase 2 trial for ovarian cancer, with results expected in the second half of 2026.
• The company has received significant milestone payments, including $90 million from a collaboration with Gilead and $30 million from AstraZeneca.
• Compugen’s cash balance of $103 million at the end of last year is expected to sustain operations until 2027.
• The company is focused on differentiating its TIGIT antibody, COM902, through ongoing phase 3 trials with AstraZeneca.
• Compugen’s partnership with Ultima Genomics aims to enhance its target discovery capabilities.

Financial Results

• Compugen’s pipeline has the potential to generate over $1 billion in milestones and tiered royalties.
• The company ended last year with $103 million in cash, providing a financial runway into 2027.
• The GS-0321 collaboration with Gilead has yielded $90 million to date, with the potential for an additional $758 million in milestones and low double-digit royalties.
• From its collaboration with AstraZeneca on COM902, Compugen has already received $30 million and is eligible for an additional $170 million in milestones and mid-single-digit royalties.

Operational Updates

• COM701, an anti-PVRIG antibody, is set to begin a phase 2 trial in platinum-sensitive ovarian cancer in Q2 of this year, with an adaptive trial design that allows for potential accelerated approval.
• COM902, an anti-TIGIT antibody, is involved in eight ongoing phase 3 trials through a collaboration with AstraZeneca.
• The GS-0321 program, managed by Compugen, is currently in phase 1 trials under a unique collaboration with Gilead.
• Compugen is leveraging its collaboration with Ultima Genomics to enhance its computational platform for target discovery.

Future Outlook

• Results from the COM701 ovarian cancer trial are anticipated in the second half of 2026, with a goal of extending progression-free survival by three months.
• Compugen awaits results from AstraZeneca’s eight phase 3 trials involving COM902.
• The GS-0321 program will continue with dose escalation in phase 1, followed by expansion into specific indications based on data.

Q&A Highlights

• Investors are keen to see data on TIGIT assets, particularly the differences between Fc active and Fc inactive profiles.
• Compugen believes its PD-1 TIGIT bispecific will be safer for combination therapies compared to PD-1 VEGF bispecifics.
• The anti-IL-18 program is initially targeting solid tumors due to high IL-18 levels in the tumor environment.
• The company is optimistic about PVRIG blockade’s potential in treating ovarian cancer, citing a favorable safety profile.

For a detailed analysis of Compugen’s presentation, readers are encouraged to refer to the full transcript below.

Full transcript - 24th Annual Needham Virtual Healthcare Conference:

Ethan Markowski, Member of Biotech Research Team, Needham and Company: Good morning, everyone, and thank you for joining us today at Needham and Company’s twenty fourth Annual Healthcare Conference. My name is Ethan Markowski, and I’m a member of the biotech research team here at Needham. Joining me today is Aaron Ophir, chief scientific officer of Compugen, who will provide a presentation on the company followed by a q and a session with the remaining time. And with that, I’ll go ahead and turn it over to Aaron.

Aaron Ophir, Chief Scientific Officer, Compugen: Thank you, Ethan, and thank you to the Needham team for inviting us to your conference. So Compugen is a clinical stage computational discovery company that focus on the field of immuno oncology. And we use our computational platform called Unigen to identify novel drug targets in the field of immuno oncology, and the platform is validated by multiple targets, making them all the way for computational discovery up to clinical testing and also validated by our partnership with a few pharma companies, and we are our pipeline currently is eligible for up to more than a billion dollar of potential milestones and tiered royalties. So using our potential platform, we identified TIGIT around the same time Genentech did, and then a few years later, we identified PVRIG. So we have in our pipeline two fully owned assets, COM701, which is a potential first in class anti PVRIG antibody, and COM902, which is a potential best in class anti TIGIT antibody.

We also licensed the right to use our potential best in class TIGIT antibody as part of the TIGIT PD one bispecific rilvagostomy of AstraZeneca, and they are now moving in multiple phase three studies with this molecule, and they are expected an estimation of sale of more than 5,000,000,000, so potential blockbuster for this molecule derived from our TIGIT antibody. More recently, we used the our computational platform Unigen to identify another first in class potential asset and really novel way to harness cytokines for immunotherapy. So we developed an antibody that was licensed last year to Gilead called now GS zero three two one, and we’ll elaborate a bit later about this unique asset and the and the deal terms. And, also, we have some other undisclosed early stage assets, all of them derived from our computational discovery platform. We have a solid cash balance to the end of last year, $103,000,000 that should enable us to a a runway into ’27.

So let’s discuss a bit our pipeline. I’m gonna discuss first with our first in class potential asset COM701, which is a PVRIG blocker target again that we identified computationally. And as you can see, there are some movers that that follow us, some some with monoclonal, some with bispecific antibodies. But we are the first to move, and we are we have a very unique clinical program because we rely on the fact that only not only we identified PVRIG as a novel checkpoint, but we also realized that PVRIG has a very different biology from all the other checkpoints, seems to be dominating other stages of a cancer immunity cycle. And therefore, the activity we have seen for COM701 across different indication and across a few clinical trials seems to be very different from what we see from other checkpoints due to that unique biology that have the potential to enable checkpoint or PVRIG blockade in indication which are typically not responsive to other checkpoints.

One of these indications is a is ovarian cancer. So even before we started, we knew that PVRIG has a very high expression in the pathway of PVRIG, has a very dominant expression in ovarian cancer. And given its unique biology that enables has the potential to enable checkpoint blockade in less inflamed tumor types, we started by testing COM701 in monotherapy and then moved into combinations. And the signals we have seen there supported activity in platinum resistant ovarian cancer. In last line treated patients, we saw a signal that was above the signal we have seen previously from other checkpoints from what was shown for PD one or even PD one plus TIGIT.

So we think that we have a signal for PIVRAG blockade in platinum resistant ovarian cancer. Then we made a decision after publishing the results end of last year in SITC. We had few options. One of them would continue in the last nine patients, but after we substantiated the signal in the last last patients, we decided to move earlier. We have seen that in our trial, the patient who did respond had a very good durability, a very good safety profile, and we thought that this is exactly the kind of drug that could be relevant actually in earlier settings.

We’re going now into the platinum sensitive settings in which patients are have a more potent immune system because they are less exhausted with all the chemotherapies. We are taking patients who receive after second or third line, receive the platinum, chemotherapy, six cycles of platinum. These patients have now lower tumor burden. Also, the platinum, have shown to synthesize the tumor to the unique MOA of COM701. And, again, due to the fact that COM701 is well tolerated and associated with durable responses, we thought that to start with COM701 monotherapy in that settings, so it’s after the platinum as a maintenance settings.

And, obviously, also the the platinum settings is less competitive, and ADCs, are now starting to dominate the platinum resistant settings, are trying also to make a move into the earlier lines, but most of the trials are more to take these relatively toxic agents, the ADCs, to replace maybe the platinum chemotherapy. But the maintenance settings is still a a place of unmet need that we think that COM701 could fit. So looking at these trials in a bit more details, we’re talking about second or third line patients, which are still platinum sensitive. They receive the platinum after they already received BEV or PARP as maintenance in earlier settings, or maybe they’re not eligible because of different comorbidities or other reasons to receive BEV and PARP. So they receive now the platinum six cycles, and now there’s no standard of care for these patients.

So there’s really an unmet need, and this is exactly the place we’re gonna start with our adaptive trial design. So we’re starting with COM701 two two one randomization versus placebo. And, again, this is the standard of care. There is no standard of care for these patients. And we’re gonna report these results in half two twenty six.

We’re looking for a for progression free survival and, obviously, to continue to see the the safety. The the historical control for this patient population is quite solid, seems like around five point five months, so we expect to see something similar also in our internal placebo control. And we think that if we were able to extend that by three months PFS, this will be a goal to continue to other parts of this adaptive trial design. And then and, again, this we’re going to report in half two twenty six, and we’re gonna start the trial in q two this year. So after seeing the results, assuming they are positive, of course, so now we know that COM701 in monotherapy has activity in this exact settings in which it was not tested yet.

And then we have a few options. One, since we talk about the randomized trial, which is blinded, we can if the trials if the results are promising, we can continue and enroll and start thinking about accelerated approval. Or after substantiating the single agent activity, have the flexibility in the trial design to now add BEV or to add triplet of PD one and TIGIT and move forward in these settings, go earlier potentially in another trial if the results are promising. So this is for our COM701. So this is the next milestone for COM701, and we’d let’s talk a bit about COM902, a potential best in class anti TIGIT blocker.

So we all know that TIGIT started in the clinic very promising. We had the CTscape data with very promising hazard ratio in randomized study, and there are few other randomized phase two studies who show that TIGIT is active. But then in the large trials, multiple phase three trials from some of the, I would say, the the first movers in the field, Merck, Genentech, we didn’t see this phase two positive readouts translating into improved overall survival or even PFS in large trials. And and, of course, this is disappointing and and and shed a cast on all the TIGIT assets out there. But we need to remember that not all TIGIT assets are the same.

And I think one of the biggest differences is the the Fc tail of the antibody. So most of these first movers that had, the disappointing results in the phase three had an active Fc, meaning that in addition to blockade of the receptor of the TIGIT antibody, those who have an active Fc tail that could deplete cells, could manipulate cells, other things that obviously the companies who went forward with the molecule thought are beneficial, but also have potential side effects. And with the caveats of comparing across trials, we do see a strong strong trend that TIGIT antibodies with active Fc tail have worse safety profile. About the efficacy, it could also harm the efficacy by depleting the TIGIT positive cells, again, difficult to compare across trials, but the safety profile looks differentiated compared to the nonactive Fc TIGIT antibodies. And then moving into large trials, if you have patients with high rate of discontinuation that, are staying less time on on the trial, obviously, it’s more challenging to achieve this kind of, overall survival benefits in large trials.

And we always said that the right choice from our perspective for a TIGIT antibody, like the PD one antibodies, by the way, is to block it without any other manipulation. Choose a non apps non active Fc and to block TIGIT without any any undesired side effects or maybe even positive outcomes, but we thought that eventually you will not have a positive outcome by manipulating the Fc receptor, the Fc receptor binding capacity of the antibody. So we are one of the few anti companies that have such an antibody with reduced binding to Fc receptors, but also some others. And I think these are Alcus Gilead and and AstraZeneca with our own TIGIT PD one bispecific antibody, and they are now ongoing in phase three trials. So we’re really eager to see, and we think that, to see better outcomes with the TIGIT assets from our class, the non Fc active.

And, obviously, being one of the few companies out there with a non, active Fc TIGIT antibody, this kind of results in big trial and phase three trial gonna probably bring back interest into the TIGIT landscape and specifically in COM902. Now COM902 is differentiated not only by being a potential best in class and and and a NFC reduced binding antibody, but also due to our clinical strategy. So we know from others that PD-one and TIGIT probably show activity mostly in inflamed settings, in in in non small cell lung cancer, in gastric cancer. So if there’s sufficient amount of T cells and inflammation, then PD one and TIGIT blockers combined shows activity. But our strategy actually is to combine it with the pVRG blocker COM701, and what we have shown initial signals clinically and preclinically quite extensively that this unique biology of pVRG enable us to use COM701 to potentially open less inflamed tumor types to TIGIT p one blockade.

So COM701 will synthesize the tumor microenvironment, increase T cells also in less inflamed tumor types, and then the triplet could have activity also in places that until now we didn’t see activity from TIGIT and PD one blockers. So as mentioned before, AstraZeneca have licensed the rights to use COM nine zero two as part of the TIGIT PD one bispecific. They are moving now into a they have now eight phase three trials open across different indications. They reported single arm studies results that look promising, so the activity of these assets look definitely favorable compared to historical control of p d one blockers alone. And what is also interesting to mention that first of all, as mentioned before, we think that TIGIT is active and especially AstraZeneca, which also chose a nonactive Fc, we expect them to see activity even compared head to head versus frenbrot as they do in some of these trials.

But having a bispecific antibody, actually, the contribution of components that might be asked by regulators is less of a concern. So in some of the trials, let’s look at Troponlag ten for example. If you combine Rilve, the PD one TIGIT bispecific, with a Trop two ADC and compared to pembro, if the risk benefit profile of the combination of these two agents is better than pembro alone, this trial could get a win or an approval even in the absence of directly proving that is active. So as mentioned, we think that TIGIT is active, and we hope to see success upon multiple of these trials. But even if the the activity will not translate into direct ability to beat pembro, some of these trials could have a win even in the absence of that, and we are, already received $30,000,000 milestone from this collaboration.

We’re eligible for additional $170,000,000, and and more importantly, maybe, the fact that we are eligible also for mid single digit royalties. And if this drug will be indeed eventually a blockbuster, this will be a great win for Compugen as well and will enable us to further support our our pipeline using the income coming from this collaboration. Now a more recent collaboration. So using our computational plat computational platform, we’re looking at a tumor environment for resistance mechanism by tumor associated macrophages. We’re not looking for cytokines at all, but what we identified there took us into the world of cytokines.

And as you may know, cytokines, I mean, they are approved for for for thirty years, but people are really trying to manipulate them and engineer them and to overcome the inherent toxicity of cytokines and the challenging therapeutic window. And until now with not much of success, there are still clinical trials ongoing, so we’ll see. But the observation we made computationally was that in the tumor coenvironment, actually, you already have a cytokine, which is called IL-ten. It’s induced by inflammasome. It’s upregulated in tumor microenvironments.

It’s low in the periphery. So, actually, you have a cytokine there that should be active by itself, but this cytokine is naturally inhibited with another cytokine or a soluble molecule called IL-eighteen binding protein that binds this cytokine IL-eighteen and prevents its activity. So you have this complex floating tumor coenvironment in an inactive form, and we developed an antibody. No engineering. It’s a very good antibody, high affinity antibody, but just a clever approach, which is not sophisticated with all these difficulties of of engineering.

We use a blocker antibody, a drug moiety to inhibit the inhibitor to unleash IL-ten for its natural activity. And what you have shown preclinically is that this really seems to overcome many of the challenges that cytokine therapeutics has. We saw preclinically no systemic side effects. We have seen a really strong potent activity on in tumor environment, and and we and we licensed this program to Gilead. We initially received the $60,000,000 upfront on this asset.

And more recently, when we resubmitted the IND, we received additional 30,000,000. So total of 90,000,000 to date. We are now starting the the clinical trials. We started we dosed the first patient of the dose escalation in the beginning of of this year, and we’re eligible for additional 707 hundred $58,000,000 milestones and up to low double digit royalties. And what was interesting that even though Gilead licensed the assets, we’re the one moving the asset into phase one.

It’s kind of a unique collaboration approach. We have the know how. We have the capabilities. Luckily, Gilead team thought that our clinical development track is sufficient for them to trust us and to enable us to put their asset in our hands to continue into the phase one. And so we are now starting, as we said, we dosed the first patients.

The part one of the study is dose escalation in monotherapy and in combination with ZIM, the anti PD one. In the part two, we’re gonna do expansion in specific indications based on some pre some preclinical data and also, obviously, based on what we’ll see in the clinical study itself, again, in monotherapy and in combination. And and this is very good collaboration that is going on, and and and, obviously, in the next advanced stages, Gilead will take this asset. So this is the first time we’re doing this kind of collaboration, but we think there’s a very strong potential for synergism, us taking it in the first stages, and then Gilead moving in in the advanced advanced stages into the the advanced clinical trials. So and as as mentioned before, we also have other assets.

So we have a computational platform that continues to fit the pipeline. So for now, we’re not disclosing the name of these assets, but definitely, the future of Compugen will be bringing more and more assets, Licom seven zero one, Licom five zero three to clinical testing and and hopefully, eventually, to patients. And by that, I can stop and and take any questions, Ethan, you may have.

Ethan Markowski, Member of Biotech Research Team, Needham and Company: Right. Yeah. No. Thank you for the presentation, Aaron. And as a reminder, please feel free to use the question feature to submit any questions, and I’ll be keeping an eye on the live feed there.

But maybe to kick things off, and and I know you touched on it during the presentation, but Tidget, as you mentioned, obviously, a lot of, I guess, disappointment in general, more broadly. Obviously, FC Active versus FC Inactive, there’s a lot of caveats. Did you find that investors are receptive when you kind of explain where things may have went wrong and why your program’s different, or do you think that’s still a pretty high bar to hurdle right now and that they’re just waiting to see better data at this point?

Aaron Ophir, Chief Scientific Officer, Compugen: Yeah. Definitely waiting for to see data. I think that, you know, I I totally understand that. I mean, after all these phase three failures, it’s difficult to expect that this the investors will have any trust in TIGIT. I think that one needs to look into the the fine details as in many things.

And and, yes, the data does show that the Fc nonactive have better safety, have much less recontinuation rates. About efficacy, difficult to say without doing phase three trials. So I think that eventually and and it’s maybe not a coincidence that most of the companies who dropped the TIGIT antibody are are the one with the Fc active and the one with FC nonactive are still running in phase three trials. So so yes. And and, also, we need to remember that this also reflects I don’t I don’t don’t think it’s a it’s a it should, but it does reflect also in PVRIG because it is related to TIGIT in some way.

We’ve showed quite extensively it’s very different from TIGIT being active in other indication. But but eventually from the best of perspective, it’s also, I think, underestimated, but we’ll work hard to generate the data and convince.

Ethan Markowski, Member of Biotech Research Team, Needham and Company: Yeah. Yeah. That’s all I can do. So I have another sort of Tidget or, I guess, more related to your bispecific. There’s been a lot of excitement recently on a slightly different mechanism by specific, which is like the p d one, p d l one, VEGF by specifics.

I want to see, first of all, if you had any thoughts more broadly, on those on that space. And then if that you think that success maybe in that area will help bring more interest to other checkpoint bispecifics. Now I know PD one TIGIT, it’s a different mechanism, but curious if you have any thoughts there.

Aaron Ophir, Chief Scientific Officer, Compugen: So I think we have to wait and see. I mean, PD one and VEGF by principle, mechanistically, definitely have a synergistic potential. This didn’t translate enough, I would say, with Atezo and BEV, for example. And, yes, the data looks not not bad, so we’ll have to see if this translates into larger clinical trials, if the safety is indeed as good because we need to see if the p d one localization indeed is sufficient to solve the the safety issue of BEV. So so we’ll have to wait to see.

I think comparing it again across trials to the for to RILVEGO, for example. So I think the efficacy looks quite similar. Indeed, RILVEGO was not yet tested in randomized study, but just over response rate in a similar settings seems similar. So I think that eventually, mechanistically, I believe that the PD one TIGIT bispecific will be safer. So all this strategy of AstraZeneca using it as a IO backbone combining with all the ADCs, My belief that this will be a better asset for combination.

And since IO and every all oncology is going into combinations, I think that we really have to see the safety of the PDO and VEGF to be convinced they could serve the same role.

Ethan Markowski, Member of Biotech Research Team, Needham and Company: Yeah. No. That makes a lot of sense. Maybe switching over to one of the the last programs you mentioned there, your IL eighteen or anti IL eighteen. Is there any interest there in studying that program in hematological malignancies?

Aaron Ophir, Chief Scientific Officer, Compugen: So, yeah, it’s interesting question. We didn’t we didn’t look much into hematological mainly because our observation is that in the soluteur macro environment, because of inflammasome activation and cell death and all of that, there is really high amount of iodine, which is naturally there, blocked by iodine BP, and therefore the opportunity to use COM five zero three to block it, displace it, and and have this unique activity we have shown. We didn’t study it, but I’m not sure that hematological settings you will have the same type of tumor microenvironment within inflammasome and ILT naturally there. So maybe for recombinant cytokines ILT or CAR T cells secreting ILT, this could make a lot of sense. But for our program, we didn’t test it much, but we’re starting with a solid tumor due to that rationale.

Ethan Markowski, Member of Biotech Research Team, Needham and Company: Yeah. That makes sense. One one aspect, which doing a little research noticed, you have this collaboration with Ultima Genomics. I wonder if you can maybe talk a little bit about that collaboration and how you’re leveraging that technology to enhance your own platform.

Aaron Ophir, Chief Scientific Officer, Compugen: Yes. Thanks. So so maybe a few words about our platform. So we have a computational platform called Unigen in which we use different tools, including AI based tools to to identify novel drug targets. This is quite unique.

I mean, people are using AI in computational to do different aspects of r and d, but we are one of the few companies which are focusing on the first stage of really bringing new drug targets to immuno oncology. And for sure, we’re one of the very few companies which have a validated platform because we’re able to do it again and again to use computational platform, bring novel drug targets, collaborate with pharmas, take it in the clinic. So in that in that aspect, our platform is really unique in all this AI dense landscape. Mhmm. Specifically about that collaboration, so Ulta have a great sequencing technology, and we collaborate with them to to sequence patient tumor samples mostly.

And by that, we feed our own database, which is, you know, the best for all every AI is having a good, large, strong database with data of patient samples. So we are we’re using Ultima to collaborate on that. And, also, we identified it using their technology. We and this is what we published in a recent conference last month. We could identify in the single cell level specific alternative splicing events, and we are using our algorithms to employ their technology to use this kind of approaches to identify novel drug targets, so both feeding our pipeline and using the technology for specific queries looking at the novel potential targets in immuno oncology.

Ethan Markowski, Member of Biotech Research Team, Needham and Company: Very interesting. And like you said, your platform definitely drives drives the company, and and there’s a lot a lot there. Maybe one more for me, and I I will keep an eye on the live feed as well. So you are going after platinum sensitive cancer, in the maintenance maintenance setting. Obviously, refractory has been more or Prague has been kind of more of the popular place, to study drugs these days just because high unmet need.

But it sounds like there is a very significant high unmet need in the, sensitive setting as well. Why do you think more companies either aren’t going after sensitive or have been unsuccessful so far? Is it just a safety therapeutic window, or or what do you think?

Aaron Ophir, Chief Scientific Officer, Compugen: So first of all, I think that every new drug normally goes into the last line patient. This is where you start. You want to look for your signals, and this is what we did as well. But after substantiated the signal, in the last line settings, we had some patients in monotherapy, including responses responsive monotherapy and then and then the combinations. Moving into the earlier platinum sensitive made a lot of sense.

It seems that there is an unmet need. There were a few attempts, not nothing maybe not in this exact population that we identified, but there are few attempts for maintenance settings with checkpoints. I mean, also, other thought is a good idea and with not much success to date. So as mentioned, we think that PVRIG blockade should and could and should be different. This is because of the really unique biology of COM seven of PVRIG compared to all the other checkpoints, which there is a biological reasoning why it could be active in ovarian cancer in general.

And also because we’ve seen the signal. The signal seems to show differentiated activity. So having a safety profile of a checkpoint but unique activity in in in these settings should enable us or could enable us to show activity in the platinum sensitive settings. So as mentioned, the ADCs, which are dominating or starting to dominate the the the probe settings, are trying to move now also in the set the the soc, the platinum sensitive. But, again, most of the studies, and probably rightly so, we we’d still talk about the toxic agents, are aimed to replace the platinum or the the chemotherapy.

So, yeah, in the in the maintenance, some of the few attempts for PARP after PARP, BEV after PARP, and all these attempts, but there there’s definite unmet need of maintenance. And, yeah, I think that we have a really interesting potential for COM701 there.

Ethan Markowski, Member of Biotech Research Team, Needham and Company: Yeah. No. And the adaptive study design and where you could start monotherapy and then, you know, work your way to combination if necessary. I I think it’s definitely a unique way to to address the issue. I think taking a look at the chat, I don’t see any further questions at this time, but I just wanna thank you for coming on today and taking the time to talk a little bit more about Compugen.

And, yeah, thank you for attending our conference.

Aaron Ophir, Chief Scientific Officer, Compugen: Thank you.

This article was generated with the support of AI and reviewed by an editor. For more information see our T&C.