Compugen Ltd. (CGEN) Earnings Call Transcript & Summary

Ladies and gentlemen, thank you for joining us today. Welcome to the Compugen Investor Conference Call, ESMO Immuno-Oncology. [Operator Instructions] An audio webcast of this call is available in the Investors section of Compugen's website, www.cgen.com. As a reminder, today's call is being recorded. I would now like to introduce Yvonne Naughton, Head of Investor Relations and Corporate Communications.

Thank you, Miguel, and thank you all for joining us on the call today. Joining me from Compugen with the prepared remarks are Dr. Anat Cohen-Dayag, President and Chief Executive Officer; and Dr. Henry Adewoye, Chief Medical Officer. We are also delighted to be joined by Dr. Oladapo Yeku, Assistant Professor of Medicine, Harvard Medical School, and Director of Translational Research, Gynecologic Oncology Program, Massachusetts General Hospital, Boston, who was an investigator on both our dual and triple combination ovarian cancer studies; Dr. Eran Ophir, Senior Vice President, Research and Drug Discovery; and Alberto Sessa, Chief Financial Officer, who will join us for the Q&A. Before we begin, we'd like to remind you that during this call, the company may make projections or forward-looking statements regarding future events, business outlook, development efforts and their potential outcome, the company's discovery platform, anticipated progress and plans, results and time lines for our programs, financial and accounting relating matters as well as statements regarding our cash position. We wish to caution you that such statements reflect only the company's current beliefs, expectations and assumptions but actual results, performance or achievements of the company may differ materially. These statements are subject to known and unknown risks and uncertainties, and we refer you to the SEC filings for more details on these risks, including the company's most recent annual report on Form 20-F filed with the SEC on February 28, 2022. The company undertakes no obligation to update projections and forward-looking statements in the future. And now, I turn the call over to Anat.

Thank you, Yvonne. Good morning and good afternoon, everyone. We're delighted to be hosting a conference call today to discuss the data we will present at the ESMO Immunology (sic) [ Immuno-Oncology ] Conference in Geneva tomorrow and which was made available online as posters yesterday. It is our pleasure to also have Dr. Oladapo Yeku on the call with us. We're making great progress. This is the second step of meaningful clinical data that we have delivered in the fourth quarter of this year in indications where there is a tremendous unmet need and where immunotherapy has typically not been responsive or shown limited responses. We have demonstrated antitumor activity in hard-to-treat tumor types in heavily pretreated patients, including patients with liver metastases in the case of MSS-CRC and with high-grade serous adenocarcinoma histology in the case of platinum-resistant ovarian cancer. And our data suggests a COM701 mediated mechanism of action, which could lead to the responses demonstrated in our COM701 drug combination. In the platinum-resistant ovarian cancer patients, which you are going to hear more about today from Henry and Dr. Yeku, all responders to the triple blockade of PVRIG, PD-1 and TIGIT were still on treatment as of data cut-off date, where 3 of these patients had a duration of response of at least 9 months. Also, it is important to note that one of the responses on dual therapy is in a patient who previously progressed on nivolumab, and this response is associated with increased infiltration of T cells into the tumor microenvironment. What's really exciting is that the antitumor activity is supported by [ this time ]. PVRIG discovered through our computational discovery engine is uniquely expressed in early differentiated stem-like memory T cell and its ligand, PVRL2, is highly expressing dendritic cells. The blockade of this T-DC interaction with our potential first-in-class anti-PVRIG could explain the increased T cell numbers in the tumor microenvironment and subsequently, the antitumor activity we have reported in this specific patient and in the rest of the MSS-CRC and platinum-resistant ovarian cancer responding patients, who are typically hard-to-treat and not responding to checkpoint therapy. Though it is challenging to compare the efficacy in the dual and triple studies, given the small numbers of patients, the increase in response rate, combined with the duration of response, does suggest a better benefit with triple treatment, in line with our preclinical and translational clinical data. While the studies, including the biomarker and translational work, are still ongoing, the totality of the data increases our confidence in a COM701-mediated antitumor effect. We, therefore, intend to pursue further studies of COM701 combination in MSS-CRC and platinum-resistant ovarian cancer. Our planned studies are aimed at strengthening the data we have already presented, further investigate the contribution of TIGIT, while building a path to future registration studies. The first is in up to 20 patients with metastatic MSS-CRC immune checkpoint inhibitor-naive patients with up to 2 lines of prior therapy treated with the triple combination of our own anti-PVRIG, COM701 and our own anti-TIGIT, COM902 and pembrolizumab. In line with our preliminary data, we believe that unleashing the 3 pathways has the potential to turn such a cold tumor to be more responsive. We have amended the study protocol, and we expect to enroll the first patient in early 2023. There has been a lot of enthusiasm from investigators in getting involved in this study, especially considering the lack of options for patients with MSS-CRC and liver met, which makes up over 70% of the patients with metastatic MSS-CRC. We, therefore, expect enrollments to be completed in 2023. When we continue to monitor the patients ongoing on study treatment, we are designing the second study in platinum-resistant ovarian cancer immune checkpoint inhibitor naïve patients. We are considering various options, including prior lines of treatment, patient population and number of patients to enable us to pursue the fastest path to registration studies. Our focused approach and choice of indications is we believe the fastest route to next meaningful data and path to registrational studies. We expect to share initial findings by the end of 2023. Although the focus of our call today is on platinum-resistant ovarian cancer, we will also be presenting data at ESMO-IO on non-small cell lung cancer. So I just want to summarize why it is important. These data are the long-term follow-up of 7 patients treated with COM701 with and without nivolumab in our dose escalation studies previously presented at ASCO last year. What it shows is that despite failing multiple therapies, including immune checkpoint, these patients derived clinical benefit demonstrated by disease stabilization. Additionally, while this is a nonrandomized and a very small study, there is a preliminary signal of improvement in overall survivors with 10 months in patients on combination therapy and 9.5 months on COM701 monotherapy. I would like to draw your attention that this is a population that has been extensively pretreated with all having had prior treatment with immune checkpoint inhibitors. Importantly, 4 of the 7 or 57% had at least 2 prior immune checkpoint inhibitors, and all 4 patients had disease stabilization with 2 out of 4 or 50% with stable disease of more or equal to 6 months. The patients have received a median of 4 prior lines of therapy all this indicating disease profile that is challenging to treat. For context, the standard of care arm in the landmark S1800A study presented at ASCO earlier this year, reported a median overall survival of 11.6 months while having received one prior line of immune checkpoint inhibitor. We are encouraged by this preliminary signal in this heavily pretreated patient population, which again suggests a COM701 role in an additional hard-to-treat patient population. Returning to the focus of our call, during today's call, Henry will start by providing an overview of the platinum resistant ovarian cancer data to be presented at ESMO-IO. Henry will then be joined in conversation with Dr. Yeku to discuss his view of the platinum-resistant ovarian cancer landscape as well as his experiences of COM701 in combination with nivolumab with and without blockade of TIGIT. And then we will open the call for questions. With that, I will hand over to Henry.

Thank you, Anat. I am very happy to provide an overview of the data that we are presenting at ESMO-IO and delighted to be joined by Dr. Yeku. There is an urgent medical need for patients with platinum-resistant ovarian cancer who have very limited treatment options. So there was a lot of excitement from investigators reporting durable shrinking or stabilization of tumors in some of their patients who have previously progressed on all other treatment options. The investigators also highlighted improved quality of life in these patients, and Dr. Yeku will give you more insight into this shortly. Platinum resistant ovarian cancer is a hard-to-treat tumor with limited T cell infiltration, which, as you know, are required for immune checkpoint inhibitors to be effective. Historically, most immune checkpoint inhibitors have demonstrated limited activity in such tumors. Effective treatment options for patients with platinum-resistant ovarian cancer are limited. Standard of care in patients with platinum-resistant ovarian cancer includes single agent chemotherapy with response ranging from 8% to 12% with median progression-free survival of 3 to 4 months and overall survival of around 1 year. Immune checkpoint inhibitors as monotherapy and as part of immunotherapy combinations have demonstrated limited activity in this patient population. We believe that PVRIG's unique biology, different than other checkpoints, has the potential to generate different outcomes. The data published today in platinum-resistant ovarian cancer patients includes 20 patients from the dual combination cohort expansion study and 20 patients from the triple combination cohort expansion study. The baseline characteristics include a heterogeneous patient population with diverse histology. The most frequent histology was high-grade serous adenocarcinoma in 13 of the 20 patients, representing 65% of the patients in the dual combination study and 11 of 20 or 55% of the patients in the triple combination study. The patients were heavily pretreated with a median of 4 prior lines of therapy in the triple and 6 prior lines of therapy in the dual combination study. 55% of the patients in the dual, which is 11 of 20 patients; and 16 of 20 or 80% of patients in the triple combination studies had prior bevacizumab; and 12 of 20 or 60% in the dual and 6 of 20 or 30% in the triple had previous treatment with PARP inhibitors. In the dual combination study, 4 of the 20 patients or 20% of the patients had prior anti-PD-1 or PD-L1 therapy. Patients who received prior immune checkpoint inhibitors were excluded from the triple study. The objective response rate was 20% in the triple and 10% in the dual. All responses were confirmed with disease control rate of 45% in both the triple and dual, including a total of 6 patients with partial responses and 12 patients with stable disease across both studies. On the triple study, 3 out of 4 of the partial responders remain on study treatment for at least 9 months. And I am pleased to say that the median duration of response has not been reached as of the data cut-off date. Although the translational data analysis is still ongoing, we are excited to report that the responses we are seeing are supported by immune activation. One of the patients with high-grade serous adenocarcinoma, who received 7 prior lines of therapy, including nivolumab plus lucitanib, an investigational agent, had a confirmed partial response following study treatment with COM701 with nivolumab, which was supported by increased infiltration of CD8 cells into the tumor microenvironment. Translational assessment of peripheral blood, including profiling of cytokines and circulating immune cells, showed a positive pharmacodynamic activation of the immune system following triple and blockade. Greater peripheral interleukin-6 cytokine reduction was detected in the triple study 1 month after first treatment among patients with a partial response or stable disease of greater than 180 days compared to nonresponding patients suggesting immune modulation in these patients. Two of these responders had low PD-L1 score or the CPS score less than 1 and 3 prior to treatment, representing patient population less likely to respond to immune checkpoint inhibitor. Further biomarker and translational work, including PD-L1 assessments, are ongoing. Previously, we reported data for a patient with primary peritoneal cancer, a type of ovarian cancer, with PD-L1 less than 1 and representing an immune desert, who was treated with COM701 monotherapy and achieved a confirmed partial response for over 18 months, which was supported by peripheral immune activation. While the numbers of patients are still small, the totality of the data, including the disease control rate, the durability of responses and the supportive translational data is encouraging, considering that this is a hard-to-treat tumor type in which the patients enrolled in the study have been extensively pretreated. In terms of safety and tolerability, COM701, combined with nivolumab with and without the MSS anti-TIGIT, showed a favorable safety profile and was well tolerated. There were no Grade 4 or 5 treatment-related adverse events. In the triple combination study, there was one Grade 3 adverse event of AST elevation, which led to study treatment discontinuation but was deemed not serious by the investigator. In the dual combination group, there was one Grade 3 adverse event of proteinuria, which led to treatment discontinuation, and this was assessed by the investigator as related to the study drugs. The totality of our data combined with the mechanism of action of COM701 justifies the further development of the triple blockade of the DNAM axis with COM701, COM902 and a PD-1 inhibitor in this patient population. As Anat mentioned earlier, we are moving ahead with the evaluation of COM701 combinations in patients with platinum-resistant ovarian cancer. I would like to extend our sincere thanks to the investigators, study staff, patients and their families for participating in our clinical trials. Dr. Oladapo Yeku, who joins me on this call, is Assistant Professor of Medicine, Harvard Medical School, and Director of Translational Research, Gynecologic Oncology Program at the Massachusetts General Hospital in Boston. Dr. Yeku is a top leader in the field of gynecologic oncology, first author on the dual combination study and an investigator on both dual and triple combination ovarian cancer studies. Welcome, Dr. Yeku.

Thank you so much for having me.

I will start with my first question. Will you please briefly describe the current treatment landscape for patients with platinum-resistant ovarian cancer?

Absolutely. It will be my pleasure. I think the first thing to understand is that platinum resistance is defined by patients who have recurrence or progression of disease 6 or more months after their last carboplatin infusion or cisplatin infusion. For this patient, what this means practically is that we can no longer use carboplatin as a treatment option. And in that scenario, that leaves us with a handful of chemotherapeutic drugs that either affects our cycle or affects the way the cells divide or undergo metabolism. Some of these just include paclitaxel or Taxol at this time given once a week. Now bear in mind, nearly all of our patients would have already received Taxol in the front line setting. So many of them are coming in with neuropathy and facing Taxol challenge again. For some other patients, we have other chemotherapy drugs, such as liposomal doxorubicin or Doxil; we have topotecan, which is sometimes given daily or sometimes weekly, depending on the side effects that we are trying to avoid; and then we have other drugs that are used way less often because of lack of efficacy, including drugs like pemetrexed and cyclophosphamide and etoposide. Now when I sit down with patients to counsel them in the clinic when faced with these options, the fundamental thing that we talk about is that platinum resistance represents a multidrug resistant state in these tumors. And because of this, many of these chemotherapy options that I outlined typically have response rates of less than 20%. Most of them are less than 10%. And because of the side effects that patients are already coming into these treatments with, oftentimes we cannot give many, many cycles. Many patients will take 3, maybe 6 cycles before they either have to discontinue because of toxicity or the tumor progresses, again because it represents a multidrug-resistant state. To sometimes mitigate this, we add an anti-angiogenic drug called bevacizumab or Avastin to the chemotherapy. This helps us improve the response rates by a little bit. But more importantly, it gives us something that we can continue in maintenance after these patients can no longer tolerate chemotherapy. So to summarize, the landscape currently for platinum-resistant ovarian cancer, bar one new -- one recently approved antibodies or conjugate, typically involves serial treatment with chemotherapeutic agents with or without bevacizumab, all with very modest rates of response and durability.

Thank you, Dr. Yeku. So considering the fact that there are many challenges in the treatment of these patients, what are some of the bothersome toxicities that your patients point out to you?

So the biggest one is peripheral neuropathy because, as I mentioned before, when we treat these patients initially, all of them are treated with the intent to cure. So we give them full doses of paclitaxel in combination with carboplatin. The problem with that is that a lot of these patients will develop residual and persistent neuropathy that is numbness and tingling in the hands and feet that can sometimes impair activities of daily living. That said, other side effects that we worry about are fatigue. A lot of our patients will report persistent fatigue. Some of this is due to residual effects of the chemotherapy that they last had. And some of this is, of course, due to the underlying cancer that ails them. Now to avoid this, we start to look at other chemotherapy drugs such as liposomal doxorubicin or topotecan. Now although liposomal doxorubicin, one of our other go-to-choices, does not have peripheral neuropathy as a side effect, it has very severe skin toxicities, and the way I describe it to patient is that anything that has a surface is a fair game for this drug. So people present with what's called hand and foot syndrome, where they have peeling of the skin in the hands and the feet. Any area of the body that has a lot of friction tends to rub off. In fact, we've seen patients get admitted to the burn unit because of severe skin toxicities from liposomal doxorubicin. Patients get mouth sores and ulcers in their mouth, which limits the ability to eat. So when these patients come to clinic, and they're getting treatment that we know will not prolong their overall survival, it becomes very important for us to consider how they are living day-to-day and what their quality of life is. So when they're not able to eat, they're no longer able to drive because of neuropathy, their blood counts are low in the case of when we give drugs like gemcitabine or topotecan and patients have to come in for transfusions and they have to go back and forth, that detriment their quality of life because of the side effects of this drug becomes a very serious concern.

Thank you, Dr. Yeku, for this. So what are your key takeaways from the COM701 plus nivolumab study, of which you are the first author on that study?

For me -- I'll start with an anecdote. I don't usually like to do this, but I think it helps us humanize the work that we do. There is a current patient, I call them patients, but can call them subjects, that I am treating on the double -- on the combination arm. And this was a lady who have had multiple lines of chemotherapy, including all of the ones I mentioned to you just now and have progressed on all of them. She had no chemotherapy options left. And this was a patient who was looking forward to her daughter's wedding that was at least 8 months away. And when we discussed this trial, I told her that the benefits of this double combination agents were twofold. One, was that there were no significant side effects. And if there were some, they would think you could easily manage. And two, the dosing schedule was also something that would not take her away from her family often. So the drug in our combination was given once a month or once every 28 days. And I'm happy to report that not only was this woman able to attend her daughter's wedding, she was able to walk her daughter down the aisle, she danced, participated in the wedding, and now the discussions in the clinic are looking towards grandkids. Now these are the types of conversations based on Dr. Adewoye's recent -- when you mentioned the median overall survival of 1 year, this maybe -- is on track to beat that trajectory. And being able to do that without having her coming to the clinic once a week for treatment, without having to give her blood transfusions, without having to manage neuropathy or skin toxicity, I think for me, is a very salient point that can be missed beyond the response rate data that everyone on the call has heard about, all the duration of response data that I'm sure everyone is looking at on the call. I think that subjective quality of life benefit that disease control and that lack of side effect that bring joy to people's lives, I think, was what really stood out for me in this study.

Thank you, Dr. Yeku. So let me just transition a little bit and ask you to comment on the histology and the responses observed in this patient population?

Certainly. So the most common -- there are several histologic subtypes of ovarian cancer. And the most common one is high-grade serous carcinoma. There are a few others, low-grade serous carcinoma, clear cell, endometrioid carcinosarcoma, pure sarcomas, mixed histologies. But we've had a lot of challenges with high-grade serous particularly because these tumors generally tend to have P53 mutations. And for those who are not aware, P53 is a tumor suppressant gene, which means that these tumor cells are fundamentally broken in a way that makes them very, very hard to treat. Now when patients are diagnosed with this particular histology, the high-grade serous carcinoma upfront, and we give them full throttle therapy, and that includes debulking surgery and at least 6 rounds of carbo and Taxol, carboplatin and paclitaxel, up to 80% of these women who have recurrence of disease within 5 years, even though they will have no evidence of disease at the end of treatment. So out of the gate, this is an aggressive tumor. And inevitably, even though all patients start out with platinum sensitive recurrent disease, after 1 or 2 treatments or 2 lines of treatments, they all inevitably become platinum resistant, and these patients will ultimately also come to disease. For this reason, many clinical trials or many trial platforms have avoided serous carcinomas, mostly because of its hard-to-treat nature. I can tell you some of the work that we do in my lab when we look at these tumors in the lab, they already show signs of multidrug resistance, even though they haven't been exposed yet to some of the drugs down the line. So fundamentally, the biology of this histology demonstrates that it's something that's very, very hard to treat and historically has not been responsive to immune checkpoint inhibitors.

Thank you, Dr. Yeku. Just a little bit, could you comment at a very high level on the triplet study also with respect to the responses observed, the improvement in quality of life we get reports from the investigators who are on the study?

Absolutely. So while we -- as a principle, we tend not to do cross-trial comparisons, each trial can still teach us a lesson in and of itself. And while we've known about the TIGIT pathway for quite some time, especially from our experiences with other types of cancers, melanoma and lung cancer being primary examples, the correct way to harness it or the best way to bring it to bear has been tricky and is likely going to be different for each type of cancer. The types of responses that we are seeing with the triplet in its flavor are very different than what we're seeing in the doublet. The depth of the response and -- both in terms of the shrinkage of the tumor, now we're talking about complete responses, not partial responses, and the duration of response, that means the ability of the immune system to stay on top of the cancer to continue to control it over a long period of time, for me, is what's very intriguing. I'm grateful that the translational data that we see is correlating with these responses. So that means that our hypothesis that simultaneous attack of these 3 different checkpoints would give us a synergistic response over the doublet seems to be coming to bear. Now obviously, we have to see the totality of the translational data and see more patients on these types of studies. But that type of research is very helpful because that tells us that we're in -- we're heading down the right track.

Thank you very much, Dr. Oladapo Yeku, for your great insights. Now I will turn the call over for questions. Operator?

[Operator Instructions] The first question is from Stephen Wiley of Stifel.

This is Tuuli on for Steve. Congrats on the data. I just have a few questions related to actually mostly ovarian. So first, can you please comment on PD-L expression status of the responding patients in the doublet study? And maybe also another question on -- like biomarker data. I understand the biomarker and translational studies are ongoing at the moment. But is there anything that actually informs us of potential benefit of adding TIGIT to this regimen? And maybe next one would be related to next steps in the ovarian cancer patients. So basically, what would be the next data disclosure of these responding patients? And lastly, yes, I have one follow-up.

So maybe, Eran, you start -- yes, Eran, you start with this, and then I'll go to the next step.

Sure. So as mentioned, the transactional work is definitely ongoing. So as discussed, we have data for 2 of the responding patients in the triplet, one of them with 100% reduction in target tumor lesion was PD-L10. We also reported in the past responded to monotherapy with PD-L10. But directly for your question on the PD-L1 levels on the other patients, this is ongoing. Regarding other translational signal, so in both studies, we see activation in peripheral blood. Overall, we do see more significant important activation also across our study, not only in these ones, in the triplet, in the peripheral blood. Also in this study, we saw an interesting reduction in IL-6 in the responding patient versus nonresponding in the triplet study, not in the doublet. So this may reflect thus having more responders on the triplet or that adding TIGITs may modulate myeloid cells because that is exclusively derived from myeloid cells but this is a speculation at this stage.

And for the next steps, while we're continuing to monitoring the patients that are on study, we're designing the next study. It will be done in platinum-resistant ovarian cancer and we'll do it in a population that is checkpoint naïve. But at this point in time, we're not sharing the design. We would share it probably in early 2023. And we're taking different parameters into consideration. Obviously, we want to strengthen or gain more confidence in the signal in the patient population that we need to -- that is reasonable for us to treat, the combination as we discuss the TIGIT contribution. And all of these are taking -- are being taken into consideration at this point in time and we'll share the design probably early 2023. With respect to the guidance of the additional data from the current patients that are ongoing on the study, we don't give the guidance at this point in time. But obviously, when we'll have more information to share, we'll share then.

The next question is from Mark Breidenbach of Oppenheimer.

Congrats on these data also. Just a few for me. I guess, just to be clear, Anat. Should we be expecting the next trial in ovarian to implement a biomarker patient selection strategy, whether it's looking at PVR, PVRL2 expression or PD-L1? Is that reasonable to expect? Or would you go for still a broad population? Second question, maybe for Dr. Yeku is -- I'm wondering whether or not he sees any reason that it would make sense to potentially layer some of these checkpoint antibodies on top of second-line single-agent chemotherapies that are currently used in platinum-resistant patients, if there's any potential advantage for combining chemo with checkpoints as we've seen in some of the other solid tumor types and across the TIGIT landscape. And a third question for me, I hope you're writing these down. I'm just wondering why in the triplet study in particular, why a decrease in peripheral IL-6 seem to correlate with response? And if there's a mechanistic basis for that? And also if the serum interferon gamma increases that you saw kind of track with response versus nonresponders in that triplet study?

Thank you, Mark. So I'll start with the biomarker 1 for the next study and then Dr. Yeku will answer the question and then Eran will go on the -- will speak on the IL-6 and interferon gamma. So we're doing -- as you know, we're saying that all the time, we're doing a lot of translational work that is informing us on the MOA of the COM701, PC, assessment and also some insights regarding potential biomarker. And you see that we're working with -- you see the data that we disclosed for the -- for what's going on in the tumor macro environment really in the forefront of the work that is -- we believe that is being done. Having said that, I think that also we need to be realistic. Biomarker in the field of immuno-oncology, it's not really very easy to identify. I think that even today, what we have out there, the PD-L1 and MSI and TMB. And we're working on this, but we'll not delay our work. While we will do the next studies and we'll continue to explore and we're specializing also on the computational work on this front. But having said that, I believe that the next study will see us taking that step for sure, the ovarian we're still in the design, but I think that it will be safe to think about the ovarian as the study that would be done and not in a biomarker selected manner, but definitely we will define and that's what we're doing now, the patient population that we will address. And that's important for us in order to increase probability of success and really test the assets and the combination what we would think is relevant. So biomarker 1, no. And Dr. Yeku?

Yes, thank you. Very quickly, I will echo your remarks regarding biomarker selection. I'll give you a concrete example. What we've seen, even though the FDA, for example, allows you to use pembrolizumab for any patient with MSI-high or TMB-high, any solid tumor, in clinical practice, we've actually found the response rate in ovarian cancer -- first of all, they're rare. But for those patients where we treat them, they don't respond in the same way that say, for example, patients with colorectal cancer or anal cancer respond or even patients with endometrial cancer in the same gynecologic tract. So the biomarkers and what they mean and whether one is enough, so it's an ongoing area of hot debate. And I think ultimately, it may require combination of various biomarkers with different cut-offs and it might also be histology or tissue dependent. Now that segues into the chemotherapy question. We've known for a very long time that there are different chemotherapy drugs that do -- that have immunomodulating properties either directly on the tumor by increasing again damage associated patterns, activating the STING pathway or even other types of chemotherapy drugs that re-modulate or reprogram the tumor microenvironment, depleting regulatory T cells, normalizing irregular tumor vasculature. And these are things that we're actively discussing with the Compugen team because now that we're focusing on ovarian cancer as one of the tissue factor going after. We're trying to leverage everything we know about ovarian cancer, what types of chemos do what. What types of chemos are a tool to immunosuppress it and what types do you potentially synergetic are looking to leverage that very carefully in this disease. So that is certainly something that we have ongoing discussions on.

And for the third question, so I will start with interferon gamma. So having the immune activation in the periphery doesn't always translate to efficient TME immune modulation and, obviously, eventually tumor reduction. So while in some patients, for example, again, that patient who responded in monotherapy and she has really a peak of interferon gamma, which is significant more than the other patients, we don't always see that. So in general, I cannot say there's a direct correlation with the interferon gamma induction and TME modulation and eventually antitumor activity. For the IL-6, so we are not sure. It could be, again, IL-6 among others, also derived from myeloid suppressor cells. So it could be that by modulating a tumor microenvironment and we have shown in the past in other indications as well, inducing interferon gamma and others, we were able to modulate the myeloid cells sufficiently also to be less suppressive and then the cytokine which is normally mostly tumor promoting like IL-6 is being reduced or it could be again something correlative and not causative. But this is -- we'll study that probably in the future.

The next question is from Daina Graybosch of SVB Securities.

I'll jump on this trend of asking a couple questions. One, I wanted to understand the MSI status of the various patients and responders. Were the patients who were previously treated with checkpoints have MSI-high tumors? Second, I've noticed that PARP exposure is pretty low, and I wonder if you could perhaps, Dr. Yeku, you could describe why if this is pretty typical of the PARP exposure and whether any of the responders had prior PARP therapy? And third, along the questions of what's the best next study to do to better understand the signal and the contribution of TIGIT and PVRIG. I wonder -- I'm really interested, Dr. Yeku, if you had to design one study, what study would that be next?

Henry? Henry, would you like to relate to the MSI status and then maybe...

Sure. Yes. So Daina, this is being collected from the site but it would be really very unusual in this patient population to have MSI tumors that are high. That's number one. And as Dr. Yeku previously mentioned, the characteristics of the responses that one sees in MSI solid tumors, it seems not to be the same when you deal with gynecologic tumors, especially with ovarian cancer. And I think Dr. Yeku mentioned that previously. At any rate, we're going on to collect some of these which are not really standard as far as ovarian cancer is concerned. I will be able to divulge some of this analysis we get at a later time point. Dr. Yeku, is there anything else you want to add to that?

Yes. So you're correct. So the types of MSI status that we'll see in ovarian cancer are typically either patients who have Lynch syndrome, who may have had endometrial cancer before or other malignancies. And in those cases, some of those patients may not be eligible for clinical trials because of concurrent malignancies or recent malignancies. The more common type that we might see is methylation of amyloid-6 and that's a very transient sort of process. Sometimes you see it. Sometimes you don't. So universally, the rate of mismatch repair deficient ovarian cancer, especially high-grade serous ovarian cancer, is generally low. We do see it, but it's pretty low.

Thank you. So just one other comment you had, Daina, and that had to do with prior bev reception or receipts of prior PARP. I think just because there are differences between the two studies, so, for example, in the poster we have, so a little above half of the patients on the drug combination that received prior bevacizumab, and well above 80% of the patients on the triple, and for prior PARP, 60% on the dual combination and 30%. I don't really think one can make much of these. I think it just depends on the patient population that one sees at the sites, the standards. The way to really figure this out is -- one way to do is studies subsequent to these, which we use this as to either have an eligibility requiring that all patients who would have received prior bev, that's one way to do it, or to use stratification to see if there's a way to balance out what the effect will be in patient who had received prior bev or not. But you do remember that as part of the discussion we had with Dr. Yeku, we mentioned that some of the toxicities preclude patients from receiving bevacizumab. So for example, the patients have -- one of the things that -- one of the side effects that patients have with bevacizumab in addition to perforation, which occurs in about 1% to 2% of patients, is hypertension and nephropathy and so on and so forth and bleeding risk also. So this might be some of the reasons why some of these patients did not receive prior bev. With regards to PARP, yes, that I will let Dr. Yeku comment also, but I would just like to say that PARP will only be -- outside the context of the clinical trial is only in patients who have gained at least a partial response and maybe continue as maintenance therapy outside the context of the clinical trials. Dr. Yeku?

Yes. So that's correct, Dr. Adewoye. So for patients with PARP -- just to quickly review, patients who are eligible for PARPs are either patients who have a germ-line in the upfront setting, patients who had a germ-line mutation and BRCA who get it as part of maintenance, patients who have a deficiency in the homologous repair deficiency pathway, who get it in combination with bevacizumab in the upfront setting. Until very recently, patients who had homologous repair deficiency can actually get PARP by itself as monotherapy or patients with BRCA mutations can get PARP as monotherapy. There has been some FDA or some voluntary road maps on PARPs based on some recent data showing potentially worse survival in these patients. So the landscape for PARPs have been kind of shifting. So either patients get it upfront or patients get it in the platinum-sensitive recurrent setting. So somebody has a complete or partial response to carbo or something or a carbo combination and then they go on the PARP. What we sometimes find is that these patients do not respond. And you cannot give the maintenance PARP if they did not respond to the induction therapy with platinum. So we see these patients who just go from chemo to chemo to chemo simply because their tumors -- they don't fit any of the indications of which our PARP inhibitors are approved. I'll also echo the notion that different sites or different regions use PARPs differently. Some people use in a very strict FDA label. Some doctors find sort of ways to get it in there to give patients breaks in between. But I think based on when this trial started, based on the amount of PARP exposure is just about right, at least for the doublet study. Now regarding some things to think about or at least I would think about if I were designing or participating in the design of the next ovarian cancer study is the line limit. So one of the things that we know based on lots and lots of data is that when you use immune checkpoint inhibitors in later lines, and the cut-off seems to be about 3 or 4 in the platinum-resistant setting, things tend to -- you tend to have worse response and worse outcomes. That's one of the reasons why this particular study, both the doublet and the triplet, are very intriguing because these are patients in whom you would not necessarily expect a response just based on the number of prior lines they have had. If I were moving forward with the next study, I would probably limit the number of prior lines 2 to 3 in the platinum-resistant study. Everybody gets one in the upfront. Then of course, keeping an eye towards registration strategy, we would want to make sure that -- I would want to make sure that this patient had bevacizumab or intolerant of bevacizumab and they have had PARP inhibitors or were tolerant of PARP inhibitors. I think that is the population that I would want to target. And then, of course, there are other things we can use in terms of patient robustness. For these types of studies, performance status, we tend to be a little more forgiving in these later-stage studies. But as we move things upwards, we tend to be stricter with those. There are other serologic markers like serum albumin, for example, that sort of correlate with sort of patient's like overall wellbeing. That will be part of it. And depending on the registration space I'm looking at, I would design either a triplet study by itself, just like the one we have now in the appropriate space or a chemo combination, just like we were talking about before, of course, being careful as to what the follow-up study to that would be when I go up for discussions with the FDA. But certainly, adjustments in the eligibility criteria would be the first thing I look at. And this is proper course for the development of these types of drugs.

And what arms or combination, Dr. Yeku, would you recommend in such a study?

Can you say it again, what arm did I?

Yes. So I heard the setting. Thank you for all that on the setting. What specific drugs or arms would you test in this next study?

So one option would be the triplet, [ despite ] in an earlier line dual lines of therapy and do sort of a definitive study in that space. I would be -- I would look very carefully at the previous types of chemo that people have had because what the FDA might want to see is what is the benchmark chemotherapy drug -- if they ask me to a randomized study, I need to be somewhat confident that these are patients who will not benefit from weekly Taxol or something like that. If I were to do a doublet study just from 701 and an anti-PD-1, I might add chemotherapy to that and the choice, again, will depend on what I think might be immuno-modulating or immuno-activating. And there are a couple of options for that. And I would think very carefully about, again, my registration strategy would have to be a triplet because that will be chemo plus the doublet that I think will have a pretty good chance in patients who have progressed on other treatments. So for example, in the clinic, in patients who have progressed on weekly Taxol in the platinum-resistant setting have nowhere to go. You have Doxil, you have GEM, you have all these other things, but weekly Taxol is your most efficacious monotherapy agent in that space. And that's why you'll see a lot of registration trials the FDA asks people to compare the drug against either investigator choice or dose weekly Taxol. So if I test my drug in the population of patients that's post weekly Taxol, the responses for most chemotherapies in that setting is 10% to 15%, and many patients who have already had those chemotherapy options, and we're looking at nothing. So those are the sort of things that I will be looking at, both in terms of line limit or more specifically the types of the exact treatment history that these patients have had to really make sure that I am filling a need. And right now, with platinum-resistant ovarian cancer outside of -- like I mentioned before, the recent approval of mirvetuximab, there hasn't been any new approvals in the last 10-plus years. And the thing with mirvetuximab is you have to have folate receptor alpha. So this is not going to be a drug that's going to be available for everyone. So it's not really a comparator for this discussion. I hope that answers your question.

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

Yes. Thank you, operator. Thank you all for joining us today and taking the time to follow the company. I would like to give a special thanks to Dr. Yeku for joining us on the call and providing his insights on what he really likes for the patients, how new treatment options are so badly needed and how excited he is from the data we shared. For those of you attending ESMO I-O, our team will be happy to meet you. Finally, I want to thank our very dedicated Compugen team driven by our vision to transform the lives of patients with cancer and I'm more than ever very proud to be leading these patients. Thank you.

Thank you. This concludes the Compugen investor conference call, ESMO Immuno-Oncology conference call. Thank you for your participation. You may go ahead and disconnect.