Compugen at AI Drug Discovery Conference: Strategic Partnerships Drive Progress

On Wednesday, 02 April 2025, Compugen Ltd. (NASDAQ: CGEN) presented at the H.C. Wainwright 2nd Annual Artificial Intelligence Based Drug Discovery & Development Conference. The company outlined its strategic initiatives in immuno-oncology, highlighting both opportunities and challenges. Compugen’s focus on artificial intelligence and machine learning for novel drug target identification was a central theme. The conference revealed significant partnerships with AstraZeneca and Gilead, showcasing both financial stability and potential risks.

Key Takeaways

• Compugen uses its AI platform, Unigen, to identify novel drug targets in immuno-oncology.
• Strategic partnerships with AstraZeneca and Gilead are central to advancing its pipeline.
• The company maintains a strong cash position of $103 million, expected to last into 2027.
• Upcoming milestones include phase 3 trials with AstraZeneca and a phase 1 trial with Gilead.
• Compugen’s COM701 program targets ovarian cancer, with interim data expected in 2026.

Financial Results

• Cash Position:

- Ended 2024 with $103 million.

- Expected to support operations into 2027.

• AstraZeneca Partnership:

- Total deal value of $200 million, with $30 million received.

- Eligible for up to $170 million in development and commercial milestones.

- Mid-single-digit royalties on sales over $5 billion.

• Gilead Partnership:

- Total deal value of $850 million.

- Received $60 million upfront and $30 million upon IND approval.

- Eligible for tiered royalties up to low double digits.

Operational Updates

• TIGIT Program (AstraZeneca Partnership):

- Phase 3 trials underway for rilvegostomig, targeting lung and gastrointestinal cancers.

- Early data expected in 2025.

• GS-0321 Program (Gilead Partnership):

- Compugen leads the phase 1 trial, focusing on solid tumors.

- Expansion into specific indications planned.

• COM701 Program:

- Trial in platinum-sensitive ovarian cancer to start in Q2 2025.

- Interim data expected in the second half of 2026.

Future Outlook

• AstraZeneca:

- Phase 3 data for rilvegostomig expected by 2026-2027.

- Potential to become a new immuno-oncology backbone.

• Gilead:

- Gilead to lead later stages of the GS-0321 program.

- Focus on multiple tumor types based on phase 1 signals.

• COM701:

- Interim data in 2026 will guide future trial phases or combination studies.

- Aims for a progression-free survival improvement of 3 months over placebo.

Q&A Highlights

• TIGIT Differentiation:

- COM902 is non-Fc-active, potentially offering a better safety profile.

- AstraZeneca selected COM902 for its high affinity.

• GS-0321 Mechanism of Action:

- Targets IL-10 in the tumor microenvironment without systemic toxicities.

• COM701 Ovarian Cancer Strategy:

- Focus on maintenance therapy post-chemotherapy.

- Inhibits stem-like memory T cells to enhance antitumor immunity.

Compugen’s strategic partnerships and AI-driven platform position it well for future growth. For further details, please refer to the full conference call transcript.

Full transcript - H.C. Wainwright 2nd Annual Artificial Intelligence Based Drug Discovery & Development Conference:

Ajay: Greetings, and thanks for joining us to have a conversation with Iran Ofer, CSO of Compugen. Compugen is a clinical stage biotech company that uses AI and ML to design first in class IO therapeutics. The company has established two partnerships with AstraZeneca and Gilead. So to discuss the company’s development strategy in 2025 and beyond, I welcome Iran to this fireside chat. Iran, thank you very much for joining us and having a conversation with with our audience today.

Iran Ofer, CSO, Compugen: Hi, Ajay. Thank you for having us.

Ajay: So so to to start off, can you give us an overview of Compugen Compugen’s history? I mean, and what’s the area of focus for the company at this time?

Iran Ofer, CSO, Compugen: Sure. So Compugen is a clinical stage company focusing in immuno oncology. And then as you mentioned, we’re actually pioneer in computational discovery of targets. So we use our computational platform to identify novel drug targets in the field of immuno oncology. And we’re doing it for quite a while before all this AI hype arrived.

So we’re using computational tools, including machine learning and other tools to identify novel drug targets. We actually identified TIGIT around the same time Genentech did, and later on realized that to make the full potential out of our computational capabilities, we should actually develop clinical and develop capabilities to be able to take our discoveries all the way to the clinic. And this is what we do. We have now two fully owned assets in clinical testing, COM nine zero two, which blocks TIGIT, and COM seven zero one who blocks PVRIG to checkpoints. And we also license the right to use COM nine zero two as part of the bispecific ribrogstomyc of AstraZeneca, a TIGIT PD-one bispecific that they are now testing in multiple phase three studies.

And more recently, using Unigen occupational discovery platform, we identified another first in class assets, the iodine binding protein as a target for a blocker antibody. And this one was licensed to Gilead last year, and it is now in phase one that we are leading. And then we also have multiple other nondisclosed assets in the early pipeline, all of them in the field of immuno oncology, all of them derived from Unigen, our computational discovery platform.

Ajay: Very good. So on the platform itself, you know, which you call Unigen, what is what what’s what’s the model? What’s the what’s the platform? And, also, can you highlight some some of the validations that you have gained on on this model?

Iran Ofer, CSO, Compugen: Sure. So so first, you know, this is an AI conference, and and companies that people are using AI in different stages of the research and development. They use AI to optimize drugs. They’ll use AI to identify biomarkers. And what we do is quite unique because we go to the very first stage of identifying the targets themselves.

So we use AI and other computational capabilities to identify new drug targets. So, eventually, if you look at the clinical trials, phase one, phase two, phase three, most of them are around the limited number of targets. And we bring novel targets that we identified computationally, and I think this is one thing that is very unique. And and talking about how we employ specifically AI in our models, so we have this like in any AI, I would say that the center of it is a big database that we build along years of multiomics analysis of tumor microenvironment, and we use AI to build the database. We use AI tools to harmonize and QC the database.

And, obviously, we use AI tools like large language models to ask the query themselves to identify these novel drug targets. Now along these, not only we are able to discover computationally the targets, but we develop the capabilities how to validate the targets, how to identify quite early on which is a good development candidate, and then take this all the way to the clinic. And and I think that eventually the proof is in the pudding. Many people claim to do many things with AI, but our platform is validated because we’re quite of the few companies that are built able to identify multiple first in class assets, some of them validated by by deals with big pharma companies like AstraZeneca, Gilead, Bayer in the past, Bristol Myers, and and and all some of them also have some initial clinical proof of concept. So I think the uniqueness is in the fact that we’re focusing on bringing first in class drug targets and that the platform is validated.

Ajay: Very good. Thanks for that. And then, you know, as we started off discussing about Compugen, you talked about your partnerships. And, you know, one of the major partnerships has been AstraZeneca where, they are working on your TIGIT molecule. So in in general, how did you come about, you know, identifying TIGIT?

TIGIT has had a unique development history so far. So how does your TIGIT antibody differentiate itself against the the rest of them? So if you can if you can just start talking a little bit about the molecule, then maybe later we can discuss, know, the the the commercial potential and the economics of the deal.

Iran Ofer, CSO, Compugen: Sure. So as most of us know, TIGIT was a had a complex development history. It started off with some of the big pharmas which are actually validating data in phase two randomized studies. But later on, there were also some failures, especially in phase three, and people started to doubt the target. And then we need to remember that not all TIGIT antibodies are the same, mainly the issue of the Fc.

Some, especially one of the the four I think the first movers, Merck and and Genentech, that also later on had difficulties using Fc active TIGIT molecule that is associated with depletion of cells and potentially with some increased side effects that I think was part of the reasons that they had these challenges, while others like us, AstraZeneca, Arcus Gilead, which are now in phase three as well, chose a non Fc active, which is meant to block TIGIT like you do with PD-one, Keytruda, nivolumab, without depleting anything. And it seems to be that this kind of assets has a much better safety profile, less discontinuation rate, and given the fact that well, TIGIT was shown to be active in randomized studies. But like some of the other checkpoints, the magnitude of effect is probably not huge as we thought at the beginning. So having the active molecule, which is also safe to be able to combine it, makes TIGIT drugs which are non Fc active potentially best better in terms of risk benefit, and this is exactly how COMNO two COMNO two was developed to block TIGIT. COMNO two also have a best in class potential.

It’s very strong, high affinity antibodies, and this is probably one of the reason AstraZeneca chose this specific TIGIT molecule be incorporated as part of their TIGIT PD one bispecific, and they already shown data in in gastric, in non small cell, in, yes, in a non nonrandomized studies to this point, but the efficacy seem to be impressive and definitely better well well, taking into account comparison across trials, but better than historical data on PD ones alone and a very good safety profile. And this is probably one of the reason that AZR using this Rilve Gostomyg molecule, Rilve, as the new IO backbone. So they take this molecule, which is a good safe IO compound, and they combine it with different ADCs, chemotherapies, and now opened eight phase three trials using this molecule as a backbone and combine it with different combinations, and and we are really eager to see how this will develop.

Ajay: So in in so two questions from that. So one, you said AstraZeneca has a, you know, has a large set of clinical programs, you know, on the on the back of So what sort of indications are they going after, and what sort of data would we be seeing in the next couple years? And then on the on the on the economics itself for Compugen, you know, what what is I how much can you tell us about what the milestones payments are or when they could be coming?

Iran Ofer, CSO, Compugen: Sure. So so they’re following mostly non small cell, different trials. Some are in the early settings, some are in the late stage settings, and gastrointestinal. They are combining with either chemotherapies or ADCs from b seven h four, Trop two, Inerto. And I think what is nice to to mention that even though we’re quite certain that TIGIT is active and that Rilove is more active than pembro and other PD ones, in some of the studies, they’re actually not even comparing necessarily directly Rilove to pembro.

It’s a bispecific antibody. Nobody can ask for contribution of component. If you’re adding Rilove to Inerto and compare it to pembro plus chemo, you’re joining the force of Inerto and Rilve. And, potentially, if the treatment have a good risk benefit profile, you could get potentially approval based on that. So I think this is also something worth to mention.

They plan to show this year first glance of a nonrandomized study yet of real of a combination with ADC that probably will show us why they chose to move so aggressively in so many phase three studies. And then the phase three readout themselves probably not no later than ’26, ’20 ’7. So this is for the trials. And about the economics, so the deal was a deal of $200,000,000 total. We’re 17 now 30,000,000, and we’re eligible for up to additional hundred and 70,000,000 based on development, commercial milestones.

And what is even maybe more interesting, the fact that we are eligible for up to mid single digit royalties, given the fact that AZR predicting target of sales more than $5,000,000,000 of sales, obviously, this will will materialize, will be eligible for yearly annual income of royalties, and this will be very strong driver for Compugen to be able to continue and drive our pipeline from this income.

Ajay: Okay. So talking about income from the pipeline, you also generated you also signed a very, very large deal, I would say, with Gilead on on 03 on GS0321. What’s the molecule, and what’s the deal structure with with with Gilead?

Iran Ofer, CSO, Compugen: Yeah. So this was a really interesting one because we’re looking using Unigen, the computational platform, into tumor microenvironment to identify the resistance mechanism of tumor associated macrophages. We’re doing it for quite a while, and we developed a new algorithm. And we had this output of of suppressive mechanisms, and and one of them was this iodine binding protein. And and this is also where sometimes nonartificial intelligence, some used to call it human intelligence, is also important.

So our expertise in in immuno oncology and development enable us to pull out this one, this target out of the list. And what is unique about this target, even though we’re not looking for a cytokine program, this observation took us into the cytokine world because this target is an inhibitor of a cytokine which is naturally occurring in tumor coenvironment, a cytokine called IL-ten. So by using this antibody and IL-ten is really high in tumor coenvironment and low in the periphery. So by using this antibody, what we have shown preclinically and and the reason the target was was was attractive and pulled this relatively, as you mentioned, large deal in the preclinical stage was because we showed that potentially this target could overcome many of the challenges that other cytokine therapeutics have encountered, namely systemic toxicities, challenging therapeutic window. So what we have shown that you are modulating with a cytokine that is naturally available in a tumor co environment, we are modulating the tumor co environment in the periphery, like, completely, completely clean.

This is exactly the opposite what you see with cytokine therapeutics. So it was a preclinical asset, but was a competitive process in which Gilead eventually we signed a deal with Gilead. It was a $850,000,000 deal. We received $60,000,000 upfront. And recently, upon IND approval, we received additional 30,000,000.

What is also interesting about this deal that even though Gilead licensed the deal, we are taking it into phase one. So not only they trusted our discovery and computational capabilities for this first in class asset, they also trusted our development and clinical team to take efficiently and and and then I would say with the flexibility of of a biotech to take this drug into the phase one. And then in later stages, Gilead obviously will take over. So as mentioned, total deal value of hand 850,000,000 and tiered royalties up to a low double digit. Okay.

Ajay: In in terms of the antibody itself, you know, how did you, you know, identify this target, and what’s the mechanism of action of of of the of the of the clinical candidate?

Iran Ofer, CSO, Compugen: So the mode of action, it’s a blocker antibody. It’s a high affinity blocker antibody that displaces ILTIN binding protein away from ILTIN. So we’re actually inhibiting the inhibitor. So in the tumor environment, there’s naturally high level of this complex of ILTIN ILTIN binding protein complex together in an inactive form, and our antibody displaces ILTIN BP away from ILTIN. By that, enable ILTIN to activate very potently the tumor microenvironment and, again, without modulating the periphery.

This is really unique MOA for cytokine therapeutic in general.

Ajay: Okay. And then in in terms of the clinical indications, you know, what should what do you think are are the appropriate indications that, you know, GS zero three two one can can talk about?

Iran Ofer, CSO, Compugen: So we have shown and published that the expression profile of this ILTIN, ILTIN BP complex is quite broad in many tumors. We do think that probably one needs some basic level of inflammation or T cells and T cells to respond to this endogenous cytokine, but we think the potential is broad. We start in dose escalation in all comers and solid tumors, but then in the in the in the in the part two of expansion, probably gonna expand in specific indication based on the biology and based on the signals we’ll see across the the first part of the clinical trial.

Unidentified speaker: Great. Now I’d like to pivot to your lead program COM701. This is a PVRIG antibody. The PVRIG is also known as CD112R. Can you can you help me us understand what is this target, and why is it a good target of ovarian cancer that you are currently evaluating?

Iran Ofer, CSO, Compugen: Yes. So this is a very good question because checkpoints historically have did very poor in ovarian cancer. And PVRIG is a checkpoint, but it’s a very different one. So first, this had a very high expression, PVRIG and the ligand, PVRIG two, are very dominantly expressed in ovarian cancer. So to this was a target indication for us from day one.

But being highly expressed is definitely not enough to modulate the relatively resilient tumor coenvironment of ovarian cancer. But what we also have shown and published quite extensively is that PVLG has a very different biology and actually focusing in different stage of the cancer immunity cycle, inhibiting stem like memory T cells, which have very strong proliferative capacity. And what we have shown that if you use COM701 to block PVLG, you can really unleash antitumor immunity, and we showed it in few indications in patients which are typically not responding to immunotherapy, like patient with ovarian cancer. Now in the very last line, patients in the platinum resistant settings, in patient who failed everything, including patients who came from other clinical trials, we’re able to show some monotherapy signal. We didn’t test many patients in monotherapy, but we did show monotherapy signal both translationally with the MOA of COM701 and actual clinical long term durable response.

And then also in combination, mostly triplet combination, we saw a quite strong historical strong data compared to historical of p d one p d one plus digit. And therefore, we think that it’s not only the biology, but the clinical signals in the last line patients are supporting COM701 activity in ovarian cancer. Now looking at our results, what we identified that some of the patients had really long durable responses, excellent safety profile, but some of the patients and, unfortunately, more than fifty percent, and we eventually to have a success in platinum resistant or in any trial, we need to have improvement in median. But many of the patients relapse very fast, some of them so fast, they probably never even had the chance to respond to chemotherapy free regime like COM701 in combinations with other IO agents. So one option would be to go with a biomarker that we are working to identify a biomarker, but this is still a work in progress.

But another option will be to enrich by moving it to the earlier settings of disease, the platinum sensitive setting. And in these settings, we’re taking patients who already responded to the platinum, so they have low tumor burden. They there’s also data to show that the platinum chemotherapy sensitize the tumor microenvironment to the unique biology of COM701. We have a more active immune system. And then to put a drug which is active and safe with durable response in these maintenance settings, we thought this is the exact place where a drug like COM701 could fit.

Unidentified speaker: I see. Can you please remind us what’s the current standard of care in the maintenance setting of the platinum resensitive patient, and what do you see as the market opportunities for 07/2001?

Iran Ofer, CSO, Compugen: So we start by going into the second, third line. So the patients normally in the first line receive platinum. They receive maintenance, BEV or PAP. Unfortunately, many of them relapsed. They received maybe another maintenance, maybe depends on the HRD, BRCA status.

Eventually, when they get to the second or third line, they received the sixth cycle of platinum. They already received PAPO BEV or some of them are not eligible for many reasons not to receive PAPO BEV, and, actually, they have nothing. They all they they can do is to receive the platinum, and if they respond, to sit and wait without any maintenance or maybe some off label maintenance sometimes until they become eventually platinum resistant. So this is the exact patient population we are looking into to start with. Patient in this third or second line, not eligible for bev and PARP maintenance, and they will receive in our study COM701 monotherapy, and we compare it to a a placebo control, including very solid historical control because we know that this patient normally have a PFS of around five point five months.

Ajay: Perfect. So in in terms of interim data that we can expect in the second half twenty twenty twenty six, you know, what should we think of as a successful piece of data? And beyond that, you know, should we what what sort of development plans do you have for the seven zero one molecule?

Iran Ofer, CSO, Compugen: Yes. So this is an adaptive trial design, which allows us a lot of flexibility. So in the interim analysis in half to ’26, we plan to have results of forty patients treated with COM701, twenty patients treated with placebo, And, again, with with very solid historical control, we expect to see around six months or five point five months PFS for the control. And for us, a successful study will be around nine months or three months improvement over placebo. And then we have few options.

One, since it’s a randomized blinded study, we can if the results are really good, we can discuss with regulators and potentially even go for accelerated approval based on that study. Another option would be after understanding the magnitude of effect to start now a phase three study. And another option with this to add combinations like a bev to add the nine zero two or p d ones. So I think we have few options to move forward based on a successful readout in ’26.

Ajay: Perfect. So last couple of questions for you, Yaron. One is, as investors, what should we be looking out for either from you or from your partners, whether it is AstraZeneca or Gilead? And, also, in terms of cash runway, you know, what’s what sort of funds do you have, and how long do you think that could fund operations from here?

Iran Ofer, CSO, Compugen: Sure. So in q two twenty five, we plan to initiate this randomized adaptive platform trial of COM701. As said, we are all already ongoing now the phase one study of of COM503 or called now GS zero three two one. In ’25, also, our partner AZ plans to share early data for a real vein combination with ADC. And, again, the readout of the COM seven one study is gonna be in half two twenty six.

In addition, since the TIGIT, as mentioned before, the TIGIT sentiment out there is quite negative, but we might be influenced with any positive outcome. And there are some other positive potential positive outcomes this year from other companies, which are not our direct partners, but also could serve as a driver for our own company. And in terms of the the cash balance, we ended the ’24 with a solid cash position of $103,000,000 and with a very conservative assumptions of no royalty, no milestones, no other income, this will allow us to go into ’27 supporting all our clinical plans, supporting all our early pipeline. So in general, I think we’re in a good and solid cash position.

Ajay: Perfect. No. That’s a that’s a great place to be, especially in this financial environment.

Iran Ofer, CSO, Compugen: Yes. We’re grateful for that.

Ajay: Good luck with everything, Iran, and I hope to see you soon. If if if not in New York City, maybe in some of the conferences that we we both could be attending. And good luck to you. Thank you very much for joining us and talking to our audience today.

Iran Ofer, CSO, Compugen: Thank you so much.

Unidentified speaker: Thank you.

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