Ladies and gentlemen, thank you for standing by. Welcome to Compugen's Third Quarter 2020 Results Conference Call. At this time, all participants are in a listen-only mode. An audio webcast of this call is available in the Investors section of Compugen's Web site, www.cgen.com. As a reminder, today's call is being recorded. I would now like to introduce Elana Holzman, Compugen's Director of Investor Relations and Corporate Communications. Please go ahead.

Thank you, operator, and thank you for joining us today. With me from Compugen are Dr. Anat Cohen-Dayag, President and CEO; Dr. Henry Adewoye, Chief Medical Officer; Ari Krashin, CFO and COO; and Dr. Eran Ophir, VP, Research and Drug Discovery. Before we begin, I would like to read the following regarding forward-looking statements. During the course of this conference call, the company may make projections and other forward-looking statements regarding future events or future business outlook, the development efforts and their outcome, our discovery platform, anticipated progress and timeline of our programs, financial and accounting-related matters as well as statements regarding our cash position. We wish to caution you that such statements reflect only the company's current expectations and that actual events or results may differ materially. You are kindly referred to the risk factors and cautionary language contained in the documents the company filed with the Securities and Exchange Commission including the company's most recent Annual Report on Form 20-F filed on February 24, 2020. The company undertakes no obligation to update projections or forward-looking statements in the future. I will now turn the call over to Anat. Anat?

Thank you, Elana. Good morning and good afternoon, everyone, and welcome to our third quarter 2020 corporate and financial update. As Elana mentioned, with me today are Dr. Henry Adewoye, our Chief Medical Officer; Ari Krashin, our CFO and COO; and Dr. Eran Ophir, our VP of Research and Drug Discovery. After my corporate highlights, Henry will review the progress across our clinical programs, followed by Ari’s review of our financial statements and position. We will all be available for the Q&A session. Throughout the third quarter of 2020, we continue to advance our clinical programs and execute on our differentiated clinical development strategy with two ongoing clinical programs; COM701 and COM902. For this, we currently have four clinical studies. The COM701 monotherapy expansion study, the dose escalation study of COM701 with Bristol Myers Squibb’s Opdivo, for which we have completed enrollment and continue to collect data from patients on study treatment, the triple combination study of COM701 with Opdivo and Bristol Myers Squibb’s anti-TIGIT antibody and the COM902 monotherapy study. Our clinical strategy is designed to evaluate the role of the DNAM axis in immuno-oncology along with a new PVRIG pathway biology that we have identified in order to develop new cancer immunotherapy treatment for patient populations unresponsive to existing drugs. I'm pleased with our progress, keeping these multiple studies on track, especially in these unusual times. With encouraging data that we have shared to date and growing excitement and focus on the DNAM axis and specifically TIGIT among industry and academia in the immuno-oncology space, we are confident in our approach and in our candidates and strategic clinical path forward. Before I turn the call over to Henry to discuss these trials, I would like to highlight some recent accomplishments that support our unique path and strategy. As…

Thank you, Anat, and good day to everyone. As Anat mentioned, we'll continue to expand and are currently executing multiple studies in our clinical programs. On our COM701 program being conducted in collaboration with BMS, we are evaluating the safety and tolerability of COM701 in the therapy and in combination with nivolumab in patients with advanced solid tumors. In addition, we have a Phase 1/2 triple combination study evaluating the safety and tolerability of COM701 in combination with BMS-986207, BMS' anti-TIGIT antibody and nivolumab. On our anti-TIGIT antibody program, COM902 is being evaluated in a Phase 1 dose escalation study in patients with advanced solid malignancies. As a reminder, in April, at the oral virtual AACR conference, we reported data from the ongoing Phase 1 COM701 monotherapy and combination arm with nivolumab. This data included our monotherapy dose escalation cohorts and four or five cohorts from the dual combination arm. We and the investigators were highly encouraged with both safety and tolerability and preliminary anti-tumor activity of COM701 as monotherapy and in combination with nivolumab. We reported two confirmed partial responses on this study. In one patient with microsatellite stable platinum-resistant primary peritoneal cancer on the COM701 monotherapy arm, ongoing for 25 weeks at the time of the presentation; and in a patient with microsatellite stable colorectal cancer on the COM701 plus nivolumab arm ongoing for 44 weeks at the time of the data presentation. These confirmed partial responses with durability in tumor types typically unresponsive to immune checkpoint inhibitors were encouraging and demonstrated the potential preliminary activity of COM701 as monotherapy and in combination with nivolumab. We also have been encouraged by the durability of responses across all cohorts at the time of the presentation. We reported durable responses of at least stable disease or partial response for over…

Thank you, Henry. Good morning and good afternoon to everyone. Our financial results for the third quarter of 2020 released this morning continue to reflect the solid financial position and expense structure aimed to support our clinical programs in their various stages. R&D expenses for the third quarter of 2020 were $5.5 million compared with $4.3 million for the same period in 2019. Our R&D expenses reflect costs associated with our multiple ongoing clinical programs, which now include dose escalation of COM701 in combination with Opdivo, the expansion cohort of COM701 monotherapy, the dose escalation study for COM902 and the dose escalation study of the triple combination. Net loss for the third quarter of 2020 was $7.8 million or $0.09 per basic and diluted share compared with a net loss of $6.5 million or $0.10 per basic and diluted share for the same period in 2019. As of September 30, 2020, we had approximately $133 million in cash and cash related accounts compared with approximately $136 million as of June 30, 2020. The slight decrease in our cash balances during the third quarter is mostly attributed to approximately $8 million of operating expenses and working capital, offset by approximately $5 million of net proceeds received from exercise of warrants and exercise of employee options. Going into the fourth quarter of 2020 and beyond, we expect our ongoing R&D expenses to increase slightly as our clinical trials continue to expand and progress. Cash balance at the end of the year is expected to be approximately $124 million. In our next earnings call, we will provide a financial outlook for 2021. Thank you for joining today. And on behalf of all of us at Compugen, we hope you stay safe and healthy. And with that, we will now open the call for questions.

Thank you. Ladies and gentlemen, at this time, we will begin the question-and-answer session. [Operator Instructions]. The first question is from Stephen Willey of Stifel. Please go ahead.

Hi. This is Bonnie on for Steve Willey. I wanted to ask about your Q1 trial evaluating COM701 monotherapy. Since you’re [indiscernible] five different tumor types with four patients with tumor type, is there any color that you can provide in terms of the strategy or what evidence you'll be – that will be compelling to you when you choose which cohorts to upsize later down the line?

Yes. Thank you very much for your question. The primary objective of the study as you realized is the safety and tolerability of COM701 at the recommended dose for expansion, 20 milligrams per kilogram per week IV Q four weeks. Therefore, we're focused primarily on ensuring that we follow up on this. The tumor types that we have specified in the expansion cohort is combination of preclinical data and emerging clinical data that we previously presented first at SITC in 2019, and more recently at the AACR conference by Dr. Ryan Sullivan. So the preclinical data was guided by tumor type specified as non-small cell lung cancer, ovarian, breast, and endometrial, which we observed have high expression of PVRL2. And we added microsatellite tumor colorectal cancer as a result of the emerging data we saw first at SITC, and more recently at AACR where we disclosed we had a patient with a partial response that has been ongoing at the time of the presentation for 25 weeks. Going forward, we will continue to monitor this data. And as Anat mentioned, and I mentioned in my prepared comments also, additional data disclosures are scheduled for the first half of 2021.

Okay. Thank you. And for your biomarker analysis from patient biopsies, at what point will you be taking these patient samples? I think it'd be interesting to see, not just for patient selection in your other trials, but with all the preclinical biomarker data that you've shown, it'd be interesting to see the changes in expression of different DNAM axis members and if that correlates with response?

Yes, you're correct. So for the expansion cohort, specifically, I'm referring now to COM701 monotherapy expansion cohorts. One of the requirements for enrollment in this study is the collection of a biopsy prior to treatment and an on study biopsy also during the course of treatment. We will examine those results retrospectively and then find out if there’s correlation between expression of PVRL2 and treatment outcomes. And the treatment outcomes I'm referring to measures such as response rate and also looking at safety parameters, duration of response and the correlation with the expression itself of PVRL2, whether it's one plus, two plus, or higher. For the expansion cohort for the basket study, like I mentioned in my prepared comments, that will actually be a prospective collection. We will look at PVRL2 expression in the third cohort, which is the basket cohort, and early patients that have high expression of PVRL2 as we determine centrally will be permitted onto this study. So those are our plans. I also mentioned in my prepared comments that we will also look at PVRIG expression in addition to PVRL2 for the ongoing COM701 study.

All right. Is there any color that you would be able to provide in terms of like when you will be taking these patient samples?

Yes. The patient samples we'll be taking for the basket cohort before patients get enrolled onto this study. And for the COM701 monotherapy expansion cohort, also before they start study treatment and during the course of study treatment.

Okay. Thank you.

Your next question is from Daina Graybosch of SVB Leerink. Please go ahead.

Hi. I have two questions. One, I'm wondering if you can talk about the profile of COM701 and COM902 for Fc gamma receptor binding and engagement? And how it compares perhaps to the other version we see out there in the clinic with IgG1 that are mutated to make them Fc functionally incompetent? And I'm asking because given some of the updated data we saw at the TIGIT Summit and in recent publications, specifically from Merck and some from Agenus that the Fc gamma receptor engagement could be through myeloid or dendritic cell binding and not through the traditional recruitment of NK cells and macrophages. I'm just wondering whether the IgG4 backbone in your programs actually would engage that Fc gamma receptor on the myeloid population and if you could expect IgG4 to look different than these mutated IgG1 program? And then the second question on the new PVRIG data in the T-stem cell memory population. I know there's a recent publication that splits the T memory populations into dysfunctional and functional, and showed that the dysfunctional population had really high TIGIT and PD-1 expression. And I wonder if you've had a chance to look at PVRIG expression or replicate any of the work in that paper?

Yes. Thank you, Daina. So for the first question, so IgG4 has some binding Fc receptors, probably a bit higher than the IgG1 mutated. But overall, we expect to have similar patterns. Mainly, they're probably going to look similarly. And yes, all these are very interesting publications of Agenus and Merck that were shown in the posters in the recent years and then now in publications, indeed shown the ability or maybe even the requirement in preclinical models of TIGIT to engage Fc receptors in order to mediate this activity. And then this comes to this highly debated question how this will translate? And for example, for PD-L1, it was established as well as that PD-L1 requires Fc binding to modulate myeloid with a very similar MOA, I would say, maybe not to the fine details, but roughly the same MOA. PD-L1 was shown to require PD-L1 to modulate myeloid cells in preclinical model. And now in the clinic, nobody thinks that the PD-L1 mutated Fc binding or the Fc competent PD-1 are -- that are inferior. So basically, this is for the Fc binding question. So probably, IgG4, similar to the mutated one. But definitely, we are not sure that this is going to be translated into clinical efficacy, again, similar MOA for PD-L1 that did not translate well. And then for the second question, yes, this one publication did split the TSCM into population, one which have PD-1 and TIGIT and there might be developing to more exhausted, and this is actually a very nice publication. But actually, there are multiple publications now that shows that the PD-1 responding TSCM are key for checkpoint activity, they respond to checkpoint blockade. They are required for checkpoint blockade. This was shown in new and preclinical model. This was shown by correlation to clinical efficacy of checkpoints. So definitely, the field is evolving and is any new vertical field where you can have researches going into different directions. Overall, I still think that most of publications shown that the PD-1, TIGIT and what we now see also PVRIG expressing positive TSCM are required for checkpoint activity. Now the data that we have currently for PVRIG is correlative. We know that PVRIG has a very nice expression by TSCM, as TIGIT and PD-1. We know that PVRL2 has very nice expression and actually broader than PD-L1 and PVR on dendritic cells and [indiscernible]. And indeed, we need to translate either in patient samples or a preclinical model to transit all these interesting observations also into direct mechanistic studies. But we haven't done so yet.

Very helpful. Thank you very much.

The next question is from Mark Breidenbach of Oppenheimer. Please go ahead.

Hi. Congrats on the continuing progress and thanks for taking my questions. Just a few from me. First of all, do you think the relative dominance of the PVRIG, PVRL2 checkpoint can be directly inferred from PVRL2 and PVRIG expression profiling? I guess I'm just wondering if there are any downstream signals that we should be looking at to confirm that the inventory checkpoint is engaged as you analyze the tumor biopsies.

Yes. So in terms of signaling, the PVRIG signals through an ITIM domain, the recruitment of [indiscernible] typical to other checkpoints. It's a new checkpoint, so there's not a lot of studies. So we don't know a lot about unique signaling. We would expect to have – generally, again, increased infiltration [ph], increased T cell repair to our – in the patient samples following treatment, et cetera, due to all these recent observations in relative PVRIG/PVRL2. I don't think it's the expression level necessarily, but definitely the expression pattern is going to be interesting and we're also going to try to look at it. Maybe responding patients have not only the level of PVRL2, but maybe specific cell population, which expressed PVRL2, like dendritic cells, et cetera, are the ones very important for response. So we are going to look in all of it and correlate to clinical response eventually.

Got it. And if the expression is concentrated into these tertiary lymphoid structures, does that potentially complicate the expression profiling just due to sampling error and factors like that, or do you -- don't anticipate that wouldn't be a problem?

Well, the expression is not only the tertiary lymphoid structures. We do see very nice expression in tertiary lymphoid structures. But the expression is also on the tumor cell themselves. The expression is also in dendritic cells which are not specifically in the tertiary lymphoid structures. So basically, we don't – I don't think this is any complexity. Are we going to look at expression across all the different subtypes? Again, it's going to be that the tumor cell's level of expression is going to matter. It can be tertiary lymphoid structures, dendritic cells, different subtypes. And eventually, the end the day, also PVRIG expression of T-cell, NK cells, et cetera.

Okay, got it. And just one last one for me. I'm wondering if you can talk about any of the data that led Bayer to focus on IO naïve front-line head and neck cancer with the ILDR2 antibody. And are you anticipating any milestones from Bayer in the next 12 months with respect to this program?

So, Mark, there's not a lot that we can say about it, except, of course, is being in the public and what we, along with Bayer, that we can share with the shareholders. But basically, this is based on observations that they had in the dose escalation of their clinical study. Their analysis of then involving the last type of IO treatment and supporting preclinical data, and that was the decision we focused on IO naïve first-line head and neck. And this is – that's the only information that we can share. Obviously, this is a program that is being controlled, handled by Bayer and their expert decision-making. With respect to milestones, as always, we cannot comment on timing when we're going to pre-randomized trial, the size, but we would share and we'll share then – my thoughts obviously [indiscernible].

Okay, fair enough. Thanks for taking the questions.

The next question is from Asthika Goonewardene of Truist. Please go ahead.

Hi. This is Allen on for Asthika. Thank you for taking my question and congratulations on the progress reported. So my question is that, given all the large pharma TIGIT presentations that we've seen throughout the year at medical conferences, those presentations haven't really showed very exciting monotherapy activity, and realizing it's early and before your year-end update, I guess we're curious to know what would you like to see in your COM902 Phase 1 that would give you the confidence to proceed with the program and maybe to eventually replace the first TIGIT in the triplet study?

Yes. Thank you, Allen. So, as you know, this is the first time we're doing a clinical trial with anti-TIGIT antibody from 902. So for this Phase 1 study, the focus primarily is on safety and tolerability of COM902. And as we do the ascending dose escalation cohorts, we will keep on accruing data on the patients that enrolled. And we're all familiar with the data presented by Roche, Genentech, the data that was previously presented by Mereo, OncoMed, and more recently the data by Merck regarding the activity of anti-TIGIT antibodies as monotherapy and in combination with the PD-1 or PD-L1 inhibitor. I cannot comment at this time since we are still accruing data on the monotherapy activity of COM902 in terms of preliminary anti-tumor activity. But we're cognizant of the fact that in this space now that it appears that combination therapy is the way to go, and Compugen is uniquely positioned in that space. In my prepared comments, I mentioned we will eventually test a PD-1, PD-L1 free regimen. So we'll keep accruing the data and then observe for preliminary anti-tumor activity of the event. You also mentioned something about swapping BMS-986207 into either the triplet combination. So there is no plan at the moment – there's no plan to substitute our TIGIT antibody into the triplet combination as it’s currently being perceived now.

All right, great. Thank you.

I’ll just add or say that we don't see the competition between the two programs. And we've just expanded the study with BMS. We moved into a triplet study, and we'll continue to push this forward as much as possible. Our COM902, as Henry said, is really designed in order to test combinations that we can test alone, not as part of the collaboration, and mainly COM701 plus COM902 in a PD-1 independent regimen. And COM902 in general allows us to keep the flexibility and really control the two arms of the DNAM axis. So this is the reason why we have it. It's not in order to compete with the Bristol management.

Great. Thanks for taking the questions.

Your next question is from Reni Benjamin of JMP Securities. Please go ahead.

Hi. This is Justin Walsh on for Reni. It looks like your BMS and Bayer collaborations are moving forward nicely. I was just wondering when we could potentially see an update on your AstraZeneca bispecific program, and maybe just remind us about what parameters are public with respect to that partnership?

Sure. So basically, this partnership is – it's a license agreement for AstraZeneca to develop bispecific antibodies to one of the programs in our pipeline. We kept the rights for monotherapy and combination therapy and also some bispecific rights in AstraZeneca and got the total rights for bispecifics other than what we have. And as I said, it's a license agreement. And under the agreement, AstraZeneca is in charge of all the R&D and commercial activity. And in addition to the milestone that we got when we entered this collaboration, we're eligible for packages of milestones and royalties for each and every product that is going to be developed based on this license that AstraZeneca obtained. So, as such, it is really on AstraZeneca to report any advancement in this program. We'll obviously share when we will obtain milestone payments, but we cannot say anything more than that.

Got it. Thank you. And I guess just one more clarification on the biopsy data. So, of course, you're going to look for the PVRL2 and PVRIG expression and correlate that to response. So I was just wondering if there's anything else specifically that you'll be looking for to identify responders or get some sort of a sense of the drug efficacy.

So basically, we're going to look at multiple parameters. When we look at the pathway expression, probably also other clinical molecules, we're also going to look at the TIGIT [ph] general modulation, we're going to look at T cell sequencing, RNA cells. So we're going to try to see – look for drug MOA, hints of activity, along with potential exploratory biomarkers as well.

All right. Perfect. Thank you for taking our question.

Your next question is from Tony Butler of ROTH Capital. Please go ahead.

Thank you. Good morning. Just back to the biology a little bit, especially with the recent TIGIT Summit. So the question is, it's really one, but it has four parts. And the first is around when you look at over-expression of PVRL2, I'm curious if in fact that's in tumors, which have already seen, maybe they’re second or third line tumors, in other words, they're not front-line tumors, so the over-expression is present there. And the question is, if one were to look at a naïve patient that maybe even had breast, endometrial, ovarian cancer, would you also in those naïve patients see over-expression of PVRL2, or is that a hypothesis? That's part A. Part B is, if you have over-expression of PVRL2 on DC, what's the level of expression of PVR? Is it there at all? And then Part C is, if you have over-expression of TIGIT in Treg, would you not think that always using an anti-TIGIT antibody would be useful because this gets into this dominant question yet again? And then finally, does – and it's just straight biology, is there – there is some question about PVRL2 binding to PVRIG. Do you get any binding to TIGIT as opposed to PVRIG? In other words, there is some competition and I just don't know the kinetics behind the differences in the binding. Thanks very much.

Okay. So for the first question about the expression of PVRL2 on naïve patients, well, actually PVRL2 is expressed quite broadly. We expect it to be expressed on naïve patients, on treated patients. We look at multiple patient samples, both by IC [ph], by flow. We don't expect to see any specific differences between IO naïve or heavily preceded patients there. The expression is probably going to be there in all patient samples. In terms of the expression of PVRL2 versus PVR on DC, so there are two subtypes of disease, one of them is activated disease. These were the subtypes which was highlighted to be the ones which actually capture the antigen, [indiscernible] lymph node, activated T cells over there. So PVRL2 has a very dominant expression or activated disease, but PVR and PD-1 are also expressed in this specific – this subtype. So there is co-expression of the three ligands on activated DCs while PVRL2, it is by RNA, it’s more dominant on the activated disease compared to PVR. But then there are also other subtypes of disease, DC1, DC2, which also plays different roles in T cell priming, different publications about all of them, very important in T-cell priming. So PVRL2 is expressed also in these subtypes. And again, from what we know at this point and it's mostly RNA data by now, but multiple data sets, it seems that PVRL2 has broader expression on the other DC subtypes compared to PVR and PD-L1. Then for the third question about TIGIT on Treg. So I'm not sure if I got it right, but basically what is nice about TIGIT that it may be in contrast to some other checkpoints, a blocking antibody. One the hand, we'll stimulate effector T cell activity but it will also reduce Treg suppressive activity. So it's not a competition that whether you're going to have some checkpoints if you block them. Actually, the Treg will be more suppressive. This is not the case for TIGIT as far as we know. So the same blocker antibody you have a double benefit effect of enhancing T cell activity and reducing Treg suppressive activity. And finally, for the question of PVRL2, PVRIG, PVR, so the binding affinity of PVR to TIGIT is in the nanomolar range. The binding affinity of TIGIT to PVRL2 is in the nanomolar range. And actually we hardly even saw it when we tried to look at binding. Not to mention that when you try to look functionally, if there is a consequence of PVRL2 binding to TIGIT, we didn't see much. On the contrary, PVRIG binding to TIGIT to PVRL2 is, again, nanomolar, in terms of nanomolar range. And of course, we saw in multiple essays that PVRL2 binding to PVRIG is functional. So basically, there is some binding of PVRL2 to TIGIT. We believe this is not dominant, definitely not competing with the dominant binding of PVRIG to PVRL2.

Absolutely perfect. Thank you very much.

We have a follow-up question from Daina Graybosch of SVB Leerink. Please go ahead.

Thank you. Thanks for a follow-up. It's actually very related to the last question. I was just wondering if you had any thoughts on some of the recent report of Nectin-4 being another ligand for TIGIT. And if you think that's at all relevant and you're thinking about that in your development of your TIGIT combination?

So, yes, it's an interesting report. I think it's one publication showing that Nectin-4 also binds TIGIT. According to that publication, actually, the affinity might be relevant. At this point, we didn't look much into this. And certainly, we didn't disclose anything, but we are not sure if it's very relevant. The binding site of different Nectin is usually similar. So I would expect, but we didn't test, that the normal TIGIT antibodies will block also the binding to Nectin-4. But again, we didn't test it and we didn't look too much into this.

This concludes our Q&A session. I would now like to turn the call back to Compugen's President and CEO, Dr. Cohen-Dayag. Would you like to make your concluding statement?

Thank you. We are very pleased with the steady progress and execution across our clinical program, ensuring all on track to our expected data readouts. In addition, starting our triple combination study in parallel to these other studies was an incredibly important milestone, advancing our broader science-driven clinical approach to a direct evaluation of our underlying hypothesis. We're excited about what's to come. In 2021, we expect our data readouts to provide additional insights into the role of PVRIG in cancer immunology. We also look forward to sharing additional guidance on the triple combination study and development plans for COM902 beyond the monotherapy dose escalation study. Thank you again for joining us today and your interest in Compugen. Stay safe and healthy. Thank you.

Thank you. This concludes the Compugen Ltd. third quarter 2020 financial results conference call. Thank you for your participation. You may go ahead and disconnect.