Compugen Ltd. (CGEN) Earnings Call Transcript & Summary

Hello, everybody. Welcome to our fireside chat with Compugen. Thank you to the company's management team for joining us. It's a quick chat today. So I'd love to jump straight into discussion here.

Computational drug discovery has a bad reputation, a lot of times. Tell us a little bit about how Compugen's approach is different?

Yes. Thank you, Jonathan. And I can see why you're saying that. I think that in the last 2 years, it is gaining more momentum and attention. Compugen has a long history with respect to computational discovery. Actually, we've been for 20 years, computational discovery company, and then we moved to use our own computational predictions for the discovery of drug targets for the development of first-in-class drugs. The way we discover new drug targets is by -- based on this 20 years of experience that we built in the company in analyzing biological data sources, multiple type of data sources, developing computer systems, tools and algorithms to analyze multi-omics data. Our technology is not limited to a specific type of data or source, it is actually flexible and we're tailoring the solution to this specific problem. Today, Compugen is in a unique situation with respect to its computational discovery capabilities. First, we have 3 drug targets that we predicted computationally that went through successful preclinical studies to the clinic. And now one of the drug targets is already clinically validated, by the way, by others as in the clinic and another drug target has initial clinical proof of concept. So that's one. And the second thing that we're distinguished on this front, on top of the proof-of-concept of the computational engine, is the fact that we're not only in silicon discovery company, computational discovery company, but we also have the drug development experience and expertise and this is integrated into the computational discovery process. And that matters.

So you talked about using omics approaches to identify computational [indiscernible] by novel targets we're also hosting companies at this conference, like Nimbus or Schrodinger, who are involved in computational drug design, is your strategy necessarily an opposition to a strategy like this?

It's not an opposition. If we put aside the technology that is being used, I think that the 2 companies that you mentioned are actually companies that are dealing with the drug design of small molecules and Compugen is focusing on the field of drug target discovery. So we're discovering new drug targets and then we develop drugs to address these drug targets, and that's different.

Makes sense. What kind of targets are best served by this sort of strategy? What sort of targets are you able to pick out? Is there a specific profile of a novel target that your engine is able to discover?

So I don't think that there is a limitation on the type of modality that would fit the drug targets that we're discovering. Specifically, Compugen is focused on discovering drug targets for antibody therapeutics and specifically we focused ourselves in the field of cancer immunotherapy. But the technology is not limited for the discovery of drug targets for antibody therapeutics. It could be applied also for drug targets for small molecules.

Makes sense. Let's jump into the programs that you actually have in clinical development right now rather than talking broadly about the platform. The lead program here is a target called PVRIG, which I think is a novel target. Most people won't have heard of it. But it is -- in keeping with that strategy of finding brand-new targets that aren't very well described in the literature already. What role does this target play in vivo under natural conditions? And where is it most relevant in oncology?

So I'll give a big -- a brief background, and then I'll ask Eran, which is heading the research and drug discovery at Compugen to add a little bit more. But just to set the framework of the discussion, yes, indeed, we discovered PVRIG by computer predictions. And actually, what we identified that on top of identifying a new negative costimulatory receptor, on T cells and MK cells, we actually identified that it is working in parallel in complement to TIGIT, another immune checkpoint that is now in the headlines in the biotech industry. But we also discovered TIGIT in 2009, through our computational discovery capabilities, and we sent it to publication back then at the time that it was published by others. So basically, we think that these are 2 -- PVRIG and TIGIT, are 2 parallel and complementary pathways that are modulating the immune system response against the cancer. And we also identified through our preclinical studies -- but this is also supported in the last 2 years by the literature, that the PD-1 pathway that is targeted today by multiple PD-1, PD-L1 inhibitors, this pathway is also working through a molecular intersection with a protein that is in the excess of TIGIT and PVRIG. In short, we actually built the basis underlying a hypothesis that addressing nonresponsive cancer patients, those that are not responsive to PD-1 blockers, addressing this patient may be benefited by blocking additional 2 pathways or one of these 2, PVRIG and TIGIT. And this is the basis of the triple pathway hypothesis that we have in the company. And we're now translating the scientific understanding of these 3 pathway story into clinical studies, which we can discuss later. But I'll let Eran focus on what PVRIG is doing in vivo and how we translate it to -- and how it is translated to clinical activity.

Yes. So basically, what is kind of unique about PVRIG -- and as you mentioned, there's not a lot of publication around it because we are -- I mean, there are not many publications, and most of them are by us because we discovered it and announced it only few years ago. But basically, what's interesting about this target is that the ligand pivotal 2 is expressed both in hot and cold tumors in PD-1 high -- in PD-1 low indications. And there are certain aspects of PVRIG pathway that we are just starting to unveil. The expression of PVRIG itself by stem-like memory T cells, which are very important, cells that can respond to checkpoint blockade and actually repopulate the tumors with new emerging prime T cells. And the ligand itself, in addition to its expression in PD-L1 low indications, also has very broad expression on dendritic cells. All in all, and to make it short, we start to see some interesting data that shows that maybe PVRIG is unique in its ability to prime new T cells and link it -- this with the emerging clinical data that probably Henry will describe, in an activity -- in preliminary activity in PD-L1 low indications. And this we indicate that this specific pathway has the opportunity to prime new T cells and maybe to [indiscernible] in less inflamed indications.

Great. One thing that you both spoke about there with potential for combinations with both TIGIT and with PD-1, a lot of investors are skeptical targets aren't expected to have any monotherapy activity. Is this target only interesting in the context of the TIGIT or PD-1 combo? Or is there the potential for a monotherapy interest as well?

I think that it's a good time to move to Henry, which is serving as our Chief Medical Officer that will discuss some of the data that we have for monotherapy initial antitumor activity for monotherapy for COM701 and also in combination with nivolumab.

Yes. So Jonathan, it's a good question you do ask and I'm just directors -- directed to the presentation that we had at AACR with Dr. Sullivan being the lead author. In that publication, we did describe 2 patients, one patient on the monotherapy arm with primary [indiscernible] cancer dose [indiscernible] resistant that had to be pressure response, okay? pressure response by stated criteria where rest inspection of 1.1. That patient had a response at the time of the data presentation of 25 weeks. As you know, this central platinum-resistant ovarian cancer or primary [indiscernible] cancer, typically have response rates that are very low and. And in fact, the time period from the beginning of treatment until the time they actually really progress, it's typically between 3 to 4 months, right? So to find a patient on the monotherapy arm we've, in response -- that is sustained for 25 weeks at the time of the presentation is actually very encouraging. And that points to the fact that there is a probability that COM701 has antitumor activity on its own, okay. You also remember in that presentation, we provided data on the disease control rate for patients on monotherapy arm during dose escalation. And we did quote a figure of 69% as a disease control rate meaning that number of patients with not just stable disease, but also the pressure response that I just mention. So that's very encouraging. The other part of the data that is also petulant is the combination with nivolumab. In that, we presented a patient with microsatellite stable colorectal cancer that had an assure response. Typically, these 2 tumor types or the tumor cancer/ ovarian cancer and microsatellite stable colorectal cancer, and typically only responsive to immune checkpoints. So the patient with MSS colorectal cancer had a pressure response through 44 weeks at the time of the presentation. So these are really encouraging. And then the other part of the question that I probably should allude to is when we do see combinations of immune checkpoints? People think -- typically think that you should have an increase in safety or toxicity if you combine 2 checkpoints. In the study that we presented, we did not observe any DLTs, and we did not observe any patients actually coming off as a result of the combination. So this 2 are very encouraging piece of the developed -- subscribers should be aware of. The disease control rates for the combination arm was 75%. So 69% of the monotherapy and 75% on the combination arm. This is quite interesting data and encouraging also.

Obviously, you're still in a basket trial. It's early in development. But as you say, that AACR data gave us this first hints of activity and some first real human clinical data, I think our teeth into. You've said there won't be new news on those clinical studies until '21. Could you be a little more specific? Could we see more data at, for instance, AACR in the spring? Especially the combinations with not just PD-1, but with TIGIT or the triple combination?

So the data that we aim to present in the first half of 2021 actually relates to monotherapy expansion cohorts to these 5 indications that we're focusing on based on the biology of the PVRIG pathway. And we will also present data that -- additional data for the COM701 plus nivo dose escalation. An additional cohort that we enrolled since the last data presentation and updated data. All this -- the totality of this data will be presented in the first half of 2021. We didn't state yet at which conference. This will be at a later time point.

Fair enough. Let's -- last few minutes here, I'd love to move on and talk about some of your other programs as well. The other clinical programs to talk -- the other major clinical program we talked about is TIGIT. Obviously, most biotech observers that are familiar with this target, at least in passing. As you mentioned earlier, you discovered this target in 2009 before you were a drug development outfit yourself and passed it on to others to develop in the clinic. How -- why develop your own molecule now, given that others have more programs and you're even in a clinical collaboration with Bristol using their TIGIT in the combination with PVRIG in the lead program? Why go back and make your own antibody?

So first, I'll say why we started it, and then I'll say why we maintain it. So the reason that we started the development for this collaboration, it was really when we understood that these are 2 parallel and complementary pathways. And we think that these are 2 key pathways within what we call the DNAM axis. And we wanted to make sure that we control the access to these 2 pathways. These are not drugs that you can buy -- TIGIT is not a drug that you can buy in the open market, and we didn't want to limit our options. We wanted to make sure that we can test our hypothesis as needed. Indeed, we are under the collaboration with Bristol-Myers Squibb, and we test under this collaboration combination with their own TIGIT. It makes sense for us because their TIGIT is more advanced than ours. It was already tested with nivolumab in clinical studies. So we just needed to titrate in COM701, it makes a lot of sense. That's a win-win for the 2 companies, and we're moving forward with it. But as of today, we don't see a reason why not to develop our own TIGIT. First, it keeps our flexibility as a smaller tech company. It allows us to go ahead and test also the blockade of PVRIG and TIGIT as a doublet without the addition of PD-1, and that's not part of the collaboration with the Bristol-Myers Squibb. And also today, when we see that TIGIT is a clinically validated pathway, if we will identify conditions under which it will make sense to test TIGIT plus additional -- other agents in combination, that's a path that we can pursue. We have in our hands now, a drug that is targeting a clinically validated pathway. So we keep our options.

Makes sense. Where do you fall on the ongoing debates about mAb isotype and Fc function for this target? Obviously, a lot of folks with advanced digit programs argue a ton about what the right way to hit it is. So how do you feel about it as the discoverers here?

Eran, would you like to relate to it?

So basically, our antibody has -- is a non-Fc binder. And the rationale for this -- and we don't a lot of time, so we'll take it in short. Basically, most of the mitigation for other companies to move into an Fc-active, that's coming from the mouse studies. Also, we did it in our hands, and we saw the same result, but then came the question, will it translate and for other clinical assets, it did not, and this is because the whole biology of Fc receptors is different in human and mouse. CD4 is depleting Tregs in mouse but not in human PDL-1 is working in mouse by binding Fc receptors to -- shown to be dependent. It's not the case in the clinic. So overall, learning from other clinical assets and from the mouse versus human Fc receptor biology. We decided to focus on pure blockers and not take the mouse as the main motivation to go into Fc binder. And without the risk of depletion of cells because actually, TIGIT is expressed by Tregs, but also by effector cells and Fc active antibody could also potentially deplete T cells.

Makes sense. Now, I know we're coming down to our final minutes. I do want to ask a few more detailed questions about TIGIT since it is top of people's minds. As we get into expansion cohorts and monotherapy in the Phase I, which you said we might expect in the first half of next year, should we be expecting those potential responses in monotherapy and how are you tracking the cohorts that are ongoing? What should we expect to see in the first half of the data [indiscernible]?

So the data for the first half of 2021 relates to COM701 monotherapy expansion cohorts and COM701 plus nivo dose escalation updated data. For COM902, our TIGIT inhibitor, we're now at the dose escalation stage. And we did state that during 2021, we will share data from the dose escalation study of COM902. That's what we have on the plan to date.

Okay. Makes sense. I guess then last question, moving on to your final program that you disclosed, the ILDR2, now licensed to Bayer. Can you talk about the clinical catalysts and development pathway for this molecule briefly? And give folks a sense of what to expect from Bayer's program?

So what we can say about this program, obviously, it is under the control of Bayer now. We entered into this collaboration at an early stage. We had a joint preclinical research, and now Bayer is actually in charge of the clinical studies. In terms of the -- of milestones, we -- Compugen are eligible for additional $250 million in milestone payments and mid- to high single-digit royalties. The specifics of the milestones -- and that's on top of $30 million that we got up until today. But the specifics on the milestones themselves, at what point in time we get the milestones and what's the milestone itself, that's under confidentiality with the Bayer, we cannot relate to this.

Okay. Makes sense. I think we're out of time. Obviously, there's a ton of questions we could ask about the novel targets, about TIGIT and your clinical development pathway. So to any interested clients on the line, please let us know if you've got more questions for the Compugen team, and I hope guys that the rest of your meetings during the conference go well. Thanks so much for joining.

Thank you very much.

Thank you.