Compugen Ltd. (CGEN) Earnings Call Transcript
November 18, 2020
Earnings Call Speaker Segments
Hi, everyone. Thanks so much for joining us. My name is Chris Howerton, I'm part of the Jefferies biotechnology equity research team. Very pleased to be hosting a fireside chat with next company, Compugen. We're obviously joined by the team here, but including: Anat, the CEO; Ari, the CFO; Henry, the clinical development lead; and Eran, the earlier-stage R&D development lead. So thanks for joining us, everybody.
Okay. Great. So maybe before we get started in terms of your lead programs in immuno-oncology, I'll note to the audience out here that you actually have a pretty extensive and robust drug discovery platform. So Anat or Eran, I don't know who would like to take this question, but would love to just kind of hear a little bit about the platform and how it is that you have arrived at your current portfolio of assets.
Sure. Thank you, Chris. So we're -- as you stated, we're a therapeutic discovery company and we develop drugs as well. The therapeutic discovery is actually discovering of new biological pathways, new drug targets that we do through our computational discovery capability that we developed for many years in the company. Basically, for more than 20 years, we developed computational systems, tools and algorithms to discover new drug targets and new biological pathways. Discovering new drug targets is a scientifically complex task in life sciences. And this is really requiring multidimensional approaches. Our capabilities are actually not focused on specific technology or specific data types. We have 20 years of knowledge in data sciences and multi-omics, including multi-omics approaches. And at the end of the day, we employ flexible approach, where we tailor-made the development of specific algorithms to answer a specific question. And this is all augmented by human expertise that we have in the company. At the end of the day, the bottom line is that we were able to come up with 3 clinical stage programs, all targeting completely new drug targets that we discovered. The key differentiating point for Compugen is that, on top of the fact that there is proof of concept of the platform that is very unique in the field of computational discovery in life sciences, it's also the fact that it's the nature of our company. We're not only in silico discovery company, but we're actually combining in silico discovery with the drug development expertise in the company. And this matters. When you start with the end in mind and you understand what it takes for a drug target to serve as a drug target for a clinical stage program, it matters. And you select differently your drug target from the -- you're doing the computational discovery work in a different manner. So this is what makes us unique, and with 3 drug targets that moved from computer prediction to the clinic.
Fantastic. And I mean, look, I totally share that perspective of kind of you can come up with an idea and a drug that might enter the clinic and be attractive from an earlier-stage perspective. But until you look at that end-stage utilization in that later-stage development, it's really hard to truly predict what the pitfalls are. And I think that, that's clearly an advantage of your strategy in the way you have the team structured. That's fantastic. Okay. So then maybe in terms of, from a high level, how many company or how many employees do you have in your company? And how are they kind of distributed?
So we have about 70 employees in the company. Most of the employees are in Israel and the clinical team is in the U.S. We have a business development person in Switzerland and the computational discovery person in Singapore. But most of the employees are in Israel. It's the -- it's focused on the R&D and G&A in Israel.
Excellent. Okay. So a pretty lean organization as well, that's fantastic. Okay. And so one of the things you mentioned is that you have 3 clinical stage assets. I think one of the ones that I'm personally most excited about is 701 with the PVRIG platform. And so maybe -- or pathway, pardon me. So maybe before we kind of get into where we are with that program, what is it that kind of led you to the discovery of PVRIG? Because as far as I know, there's not really anybody else that's going after this target. So what kind of brought you here? And what makes you think that this will be an important part of immunotherapy moving forward?
So this is the uniqueness of Compugen. We focus on the discovery of new biological pathways. And actually, it goes back to 2009, when we published the discovery of TIGIT. It was at the time that it was also discovered by others. But we did discover it computationally and sent this to publication with some validation work, presenting that this is a new [ checkpoint ]. We then got the -- at this point in time, by the way, we were not a drug development company. We were a discovery company and this is -- we sent it to publication. We did not start a development campaign for this program. But then when we turned to be a drug development company on top of discovering new biological pathways, we actually got back to this pathway. And we discovered -- at that point in time, we discovered PVRIG. And the initial understanding of the science behind the PVRIG pathway, in vitro studies, in vivo studies, expression profiles, et cetera, we actually understood that this is what we believe based on the data, that PVRIG pathways operating in parallel and in complement to TIGIT. And more than that actually that there is some connection to the PD-1 pathway. And this is all our story about the 3 pathway story, the 3 pathway connection of PVRIG, TIGIT and PD-1 and how we're focusing on translating the science into a clinical strategy and clinical studies and clinical data.
Got it. So maybe my understanding, and you can correct me when I have it wrong, is that the -- your view of the checkpoint inhibitor space for these 3 pathways is that they're all kind of related. And they -- if you do not sufficiently block all 3 of the pathways, that some sort of escape mechanism can eventually prevail. And so that's why you kind of see the complementary approach of all these different pathways. And frankly, maybe I'm jumping the shark here, but using the triple combination of all 3 might be the most advantageous. Do you think that, that's a reasonably fair way to characterize the overall perspective?
I think that it's fair. But I will add some additional data points and then I'll let Eran or Henry, if they want to add to this. But our hypothesis is that because we think that there is molecular intersection between these 3 pathways through DNAM, that in different patient populations and in different cancer indications, you'll get different dominance of each of this pathway. And this is why in certain tumor types and in certain patient populations, you'll need different combinations. The triplet, as we go ahead with our study today, is actually the most direct way to show that all the 3 are involved in certain tumor types. But it doesn't mean that in other cases, you will not need doublets, for example, which will be sufficient. And you maybe won't need the whole triplet combination. But definitely, the triplet is the most direct way to prove the hypothesis.
Yes. Okay. And Eran or Henry, did you have anything to add to that?
Yes. I think I would add that the 3 receptors, PVRIG, TIGIT and PD-1, are mostly co-expressed at least by T cells and then PVRIG and TIGIT by also by NK cells, and indeed, as Anat mentioned, connected all through the DNAM pathway either by competing for binding to ligands and by the phosphorylation of DNAM, et cetera. And what is interesting is actually that the ligands are different. I mean TIGIT binds maybe with higher affinity. TIGIT, in our hands actually, we didn't even see binding into PVRL2. But definitely, the binding is very low affinity and probably not functional. PVRIG, actually is the one who binds PVRL2. And the ligands also have a different expression pattern. And as Anat mentioned, in different tumor types, you can find different dominance of the different ligands and also across cell types. For example, in dendritic cells, a very important cell population which are important for priming, while all 3 ligands might be expressed in some subtypes, PVRL2 actually have broader expression across this subtype. So actually, the biology of PVRIG and PVRL2 is also different in a way in any indication.
Got it. Okay. Understood. And so I guess from one perspective, I think that there's mechanistic learning to be gained here, right, in terms of us understanding how this all can work and then perhaps in this one type of patient population, we should apply a doublet. But from maybe just a practical perspective, isn't it more efficient to just kind of go after the triplet as you say now, Anat? And is there any kind of safety or other concern as to why you wouldn't just pursue that as the primary strategy?
This is -- I have to say that this is now our focus as a company. Obviously, we're conducting additional studies, but this is our focus. As I said, in specific tumor types, it will make sense to test doublet. So we are making the making the necessary steps in order to be able to test it. So on one hand, we developed our own TIGIT in order to be able to test the combination of COM701 and COM902, which is a TIGIT/PVRIG blockade. We've done a dose-escalation study of COM701 nivo. And we showed data that we also had some initial antitumor activities. So it does make sense for us also to test additional combinations. But the current focus is the treatment.
Okay. Very good. All right. So then I guess I'm not -- I assume, Henry, but whoever would like to maybe address this, what -- maybe you could give us broad, broad strokes in terms of what your current clinical programs are. And then maybe I could ask a couple of pointed questions about those.
Sure. Glad to do that, Christopher. So at the very high level, we have the COM701 program and then we have COM701 program that's part of a triplet. We have already disclosed data more recently at AACR of induced escalation of COM701 monotherapy up to 16 patients and up to 20 milligrams per kilogram body weight dose IV Q4 weeks. We are currently enrolling in the monotherapy expansion cohort at that dose, okay, 20 milligrams per kilogram. Now the other study that's ongoing is the dose escalation in combination with nivolumab. And the data we disclosed was up to 10 milligrams per kilogram body weight dose in combination with nivolumab at 480 milligrams IV Q4 weeks. And that we'll talk about additional data disclosures next year. The other program and study that's ongoing is the triplet study. This is the study of dose escalation followed by a dose expansion. So the dose escalation is of COM701 and the fixed doses are the doses of BMS TIGIT 986207 and nivolumab. And at the conclusion of that dose escalation, there will be an expansion into 3 cohorts. And the 3 cohorts will be in -- the first cohort will be ovarian cancer with a 1:1 randomization. The second is endometrial cancer. And of course, the third is tumor types with high expression of PRVL2. This is a biomarker-driven one.
I understand.
Just to go back a little bit on the expansion cohort for COM701 monotherapy, it's those tumor types that Anat has previously talked about that have, by our preclinical data, high expression of PRVL2, so lung, ovarian, breast, endometrial. And you'll notice that we added a fifth cohort also, which is the cohort of patients with microsatellite stable colorectal cancer. And the reason for doing that is the data that we previously disclosed at SITC last year and then the data that then we have now obviously seen in the dose escalation cohort of our patients with a partial response that we reported at AACR. And of course, the partial response, we also reported that during dose escalation in a patient with primary peritoneal cancer. That's -- both of them are, of course, confirmed partial responses.
Absolutely. Okay. All right. So then -- so I think that, that was mainly 701. And I do note I have a question about the BMS program and collaboration. But where is 902, your own TIGIT molecule? Is that in the clinic as well?
That's right. Yes. So I just want to take break, so we can distinguish the 2 programs for [indiscernible]
Okay. Sorry, Henry, I apologize.
Yes. Right. So we have an ongoing dose escalation study of COM902 monotherapy that's currently enrolling patients and that's on track. And we -- and that's what we talk about additional information and possible discussions for all the programs that we have. But that's what is currently enrolling patients. And it is [indiscernible] clinical trial [indiscernible] as are all the other clinical studies that we have.
Okay. All right, very good. And so I guess as -- when you think about 701's activity in the clinic, one of the things that I think I like to see from a sell-side analyst perspective is if a molecule is going to eventually be moved -- used in combination, I still kind of want to see monotherapy activity and kind of how the drug will behave itself. So how is it that you see that, Henry? And what are some of the kind of things that you hope to learn from the monotherapy in terms of how it might behave in a combinatorial setting?
Right. So a few key points to highlight here. So the way Compugen is going about its development strategy is purely based on mechanistic rationale, so mechanism of action, the pathway expression. So it isn't just amassing patient populations and trying to figure out if the study drugs are working or not, okay? The other key point to note is that the studies that we are currently doing are Phase I -- mostly Phase I studies. The one that has a Phase II component to it is the triplet. And that's doing dose expansion. So the key primary objectives of this study is both for COM701 monotherapy or in combination with nivolumab or for COM902 or the triplet study. The primary objectives are safety and tolerability. That's it. Because as you remember, Christopher, that COM701 is a first-in-class checkpoint inhibitor, where the confidence of the company that has this molecule in clinical testing, so it is very important that we first figure out what the safety and tolerability is as monotherapy [indiscernible] I'm talking about is just not just short term, like within [indiscernible] window or intermediate term but also long term and therefore -- all the long-term toxicities that as we reported with other agents, just to be safe. Now one of the other objectives of the study is preliminary antitumor activity of COM701 in [indiscernible] or in combination with other agents. Now that it's important to be able to distinguish that. Remember that the patients that we're enrolling on to the studies that we have are patients who have exhausted all available standard-of-care therapies. They're not a homogenous population of patients. We have various tumor types that are being enrolled on to this study. Several of the patients have been exposed to checkpoint inhibitors and as part of standard of care or as an investigational agent on a clinical study. So the context of the interpretation of the readout will vary very much, depending on the tumor type, number of lines of therapies and whether patients have been exposed to checkpoints in the past or not. That being said, during dose escalation, we observed preliminary antitumor activity. And I would just like to put a number here, which is the one partial response in the patients on the monotherapy cohort in a tumor type that is typically unresponsive to checkpoints, right? And the other tumor type that is typically unresponsive to checkpoint in the combination with the microsatellite stable colorectal cancer. Now if you go back and look at, this is still talking about the preliminary antitumor activity, and go back and look at the disease control rate, so meaning partial responses or stable disease. It's 25% on the monotherapy arm and 3/4 or 75% are on the combination arm. It is pretty important to know what the [indiscernible] illustrates is that at least there was some benefit -- clinical benefit accruing to patients who were on the study as a result of [indiscernible] of COM701 monotherapy or in combination with nivolumab. So that's important to know. The other piece of information that's important to know is that the number of patients who actually either have partial response or stable disease were longer than 6 months, pretty significant on our study, about 6 of them overall. And several of these patients are even exceeded 200 days on the study. The other part that I also want to highlight is the durability, which is part of the [indiscernible]. So whether the patient has a stable disease or has a partial response, what is important to us clinicians is how long have these patients been able to hold on to this response, considering the patient population that is enrolled on to the study. And for now, both we and the investigators are encouraged by the data that see and we disclosed at AACR.
Okay. Excellent. I'm sorry, Henry. Did I cut you off there? I think there was a little just technical glitch. Everything good?
Yes. Everything good, I can hear you now.
Okay. Yes. Great. All right. Very good. And so -- and the next data cut from a 701 will be middle of next year or something like that? Is that kind of what you're thinking?
First half of 2021, yes.
Okay, so probably AACR or something like that. Okay. Fair enough. Okay. And then I think maybe before we discuss a little bit on 902, because I think it's also an interesting program, for Bristol, what is the nature of your relationship? And I guess, as a corollary to the 902, how should we be thinking about you utilizing their TIGIT in a combination strategy versus obviously having your own molecule in your portfolio?
Yes. So the collaboration with BMS is an active collaboration. You could see also the progress of the collaboration we started with the nivo supply and doublet dose escalation and then we moved to them supplying us their investigational drug, which is their TIGIT. And we now do the triplet together. So this is an active collaboration. With respect to the TIGIT question, our TIGIT versus their TIGIT, we moved ahead with the BMS on the triplet study. And as much as we can generate win-win in a collaboration, which is built on a win-win basis, we'll move ahead with this. We have no reason to switch horses and make changes. But on the other hand, we have our own TIGIT for 2 reasons. We want to make sure that we can test also PVRIG/TIGIT blockade independent of PD-1 blockade. And also as a small company, we need to make sure that we keep the flexibility. So we need to control the 2 pathways of the axis that we think is very relevant for immuno-oncology.
Yes. Okay. And so I think there's -- obviously, we heard not a few moments ago that you computationally discovered TIGIT. Obviously, the field has seized on to that and developed several therapies that address TIGIT. But I think one of the unique features of your molecule is that it has an intact Fc domain. So I guess how is it that you think about the differentiation from a mechanistic perspective, particularly with that particular component?
Eran, do you want to take it?
Yes. So basically, you can split this answer into 3 parts. First is why did people choose the Fc binding isotype to begin with? The second question is that, will it translate to clinical, which is obviously an open question. But can we discuss what we learned from the past? And the third part is why did we choose the non-Fc binding? So most of the motivation to choose an isotype that actually bind the Fc receptors, as Genentech and Merck did, was coming from the mouse studies, in which they infused an antibody who binds Fc receptors, either by Treg depletion or myeloid modulation or some -- different company, different mechanism described. You can indeed see better activity in the mouse settings. Now to the second part, how can this -- how will this translate? And basically, all Fc receptors system in the mouse is very different. I mean mouse is a mouse, human is human. But the Fc system specifically is completely different in human and mouse, the identity of the receptors, the cells expressing it. And if we look at how well it translated in the past, actually did not really. So CTLA-4 in the mouse, very strong Treg depletion and we don't see that in human. PD-L1 was shown to require Fc binding by Jeff Ravetch, a pioneer and leader in Fc receptor biology. And in the clinic, actually this probably nobody thinks that Viventia is better than other PD-1 antibodies. So from what we know until now, the translation from mouse to human did not work today in binding selection of isotypes. And then for the third part, what do you lose with having an antibody with an Fc binding? And then if you have Fc binding and you have any type of depletion, TIGIT expression on Tregs is overlapping with the expression on CD8 cells, the same cells eventually you want to unleash, right? So if you're going to deplete anything, you also have a risk of depleting the effector T cells and NK cells. So we chose to focus on pure blockers like PD-1 blockers, blocked interaction on [ TIGIT and PVR ] without any risk of depletion and just enhancing TNK cell activity. And this was our decision and others as well. Arcus are progressing now with AstraZeneca in Phase III trial as well. So we'll wait and see some more data from them.
Yes. Well, I mean, Eran, to your point, we've cured a lot of mice in terms of -- over the years. So I think it's definitely not the same. And I think point well taken in terms of certainly the nuances between the two. But the -- so basically, what I heard you say was that there's better activity observed in mouse models. And you think that, that could be a spurious observation. And one of the concerns that you have is that not only can the intact Fc domain deplete the regulatory cells, but they can also deplete the cytotoxic T cells as well. Is that right?
Exactly, since the expression overlap, I mean, yes.
Yes. Okay. All right. Very good. Okay. So maybe just in the last few minutes that we have together, what is the -- from a balance sheet or a corporate perspective, I guess, Ari, where is your -- hello, by the way. And where your finances at? And where have you said that, that should provide you runway to?
Absolutely. So Chris, I guess that it's -- I think it's adequate to kind of mention that when we built the budget for 2020, it was a different Compugen about a year ago, and the run rate kind of the budget was built to match the size of the company and the cash that we had back then. And therefore, it was roughly sub-$30 million. Going forward, obviously, it's a different Compugen today than it was a year ago. I think the industry and the market and the competition is different today. Clearly, we need to take advantage of the position that we're currently at and advance as fast as we can, both with our COM701 and COM902. So I think it's safe to assume that going forward, run rate will probably increase, hard to say yet by how much. But I guess we need to take aggressive approach and extend the trials by maybe additional combination or additional indication. But even having said that, we had about north of $130 million at the end of September. So all in all, I think it's safe to say that we'll have sufficient cash at least to 2024. But we'll see, as we make progress, and we'll decide what we do with the clinical trials. We are very focused and we are committed to move fast.
Okay. So essentially, last reporting, you had $130 million or so on the balance sheet.
[ $133 million ].
Obviously, expenses -- oh, excuse me, sorry. And in the past, the expenses were lower, you expect it to ramp up. You see it getting into 2024. But obviously, if there is exciting data or you having a reason to accelerate the clinical development, you feel enabled and eager to do so. Is that kind of a fair perspective?
Absolutely.
Okay. Well, very good. So I think you definitely have some amazingly exciting programs that will clearly have the opportunity to serve a wide variety of cancer patients and some exciting updates over the next year or so. It sounds like the first should be the first half of next year, which we'll certainly look forward to potentially at AACR, we'll have to see. And thank you all for joining us today, really appreciate it.
Thank you.
Thank you, Chris.
Okay.
Read the full transcript via the API
You're viewing the first half of this call. Get the complete Compugen Ltd. transcript - plus 252,000+ transcripts from 12,000+ companies, speaker segments and full-text search - through the EarningsAPI REST API or hosted MCP server.
Get an API key View API docs →For developers and AI pipelines
Programmatic access to Compugen Ltd. earnings transcripts and 252,000+ others is available through the
EarningsAPI REST API and the hosted MCP server.
Quarterly plans from $105 - full transcripts, speaker segments, full-text search,
and the /api/v1/transcripts/recent polling endpoint for ETL pipelines.