Oncolytics Biotech Inc. (ONC) Earnings Call Transcript & Summary

Good afternoon, and welcome to the Oncolytics Biotech Key Opinion Leader Webinar. [Operator Instructions] As a reminder, this call is being recorded, and a replay will be made available on the Oncolytics website following the conclusion of the event. I will now turn the call over to Jared Kelly, Chief Executive Officer of Oncolytics Biotech. Please go ahead, Jared.

Thank you very much, and good afternoon to everyone. Thank you all for joining our KOL event today. I want to especially thank our panel of experts here and all the investors and analysts who are gathered here to see -- to speak about the emerging role of pelareorep in GI cancers, including, of course, pancreatic cancer. I have the forward-looking statement slide we just passed by. Everyone should take a quick look at that and read it on your own time. The agenda today, the first half of today, is going to include presentations from our panel. We're going to start with Dr. Eggermont and go through -- again, that's about 30 minutes' worth of time. And then the back half of this event, another 30 minutes, will be a Q&A session that I will moderate. It's questions coming from our analysts and from the investor community as well as internal questions, and we'll get through as many as we can in the time that we have allotted without going over. So again, we really appreciate your interest and look forward to an engaging and insightful discussion. And at this time, I will kick it over to Dr. Eggermont to lead off our program.

So hello, everybody. My name is Alexander Eggermont, and I'm a surgical oncologist by training. I'm a Professor of Immunotherapy at University Medical Center in Utrecht, the Netherlands. And I'm on the board of the biggest, largest comprehensive cancer center in Europe, which is in Munich between the 2 universities there and where they have enormous pancreatic cancer activity. And I will give some introductional remarks about pelareorep in the pancreatic cancer program. So pelareorep, I will call it PELA from now on, is an oncolytic virus. It's a reovirus, and it is -- doesn't require any special handling. This is a great facilitating aspect. It targets cancer cells, not normal cells. And one of the biggest advantages is that it is administered IV in contrast to many other oncolytic viruses that are out there, which need intratumoral administration, which is a big stumbling block for widespread applications and certainly not an easy thing to do in peripheral clinics. It delivers double-stranded RNA into tumors. And causes tumor cell death, expression of antigens and recognition by the immune system for both angles of the immune system, the innate and adaptive immune responses. So Oncolytics has a big experience with the agent in more than 1,100 patients, and there are no safety qualms or concerns with this agent, which makes it an absolute first-in-class oncolytic virus agent. It has a clear registration path in 2 indications, strong efficacy signal in both breast and pancreatic cancer, but I will focus on pancreatic cancer. And it has Fast Track designation for both indications. So as I said, nongenetically modified, nonpathogenic reovirus, IV administration. Therefore, it targets both primary tumors, which are usually loaded with T cells and metastatic tumors, meaning that you optimize the induction of immunologic memory because of your IV administration because it can infect all tumors. And you can work both with primary and metastatic tumors. And again, it requires, as you see in yellow, no special handling. So selectively replicating cancer cells leads to accumulation of double-stranded RNA, which leads to more visibility of the tumors and it induces chemokine expression. And you need those chemokines for T cell and NK cell infiltration of your tumor. So it's one of the mechanisms by which you make a cold or semi-cold tumor hot. It induces thereby antitumor cell or immune responses, including expansion of tumor-infiltrating lymphocytes, which they have actually also shown to be circulating and can be detected in -- among the peripheral blood monocytes. So how big is the pancreatic cancer problem? It is our biggest unmet need problem. It is in terms of incidents, #10 on the list of all cancers. Number one is breast, second is prostate, third is lung, fourth is colorectal. Pancreatic is 10, but it's the second cause of cancer death. And so in 2025, you actually in the United States will have close to 68,000 new cases, but more than 52,000 in 2025, people will die of pancreatic cancer. So if that is not an unmet need, I don't know what is. So what are we doing nowadays in pancreatic -- advanced pancreatic metastatic cancer? We give chemotherapy and actually, we give combo chemotherapy. It started with mono gemcitabine, which was superior to 5FU, which was totally dismal, and that was in the old days, but still only a 1-year survival of 18% for gemcitabine versus 2% for 5-FU. That's how bad the tumor is. Now FOLFIRINOX and what you see here below, gemcitabine, nab-paclitaxel are the 2 most used standard first-line treatments. FOLFIRINOX was randomized against gemcitabine. It's quite toxic. Median overall survival, 11.1 months versus 6.4 for the gemcitabine. The combo with paclitaxel and gemcitabine versus gemcitabine at 8.5 months versus 6.7 months, but is much less toxic than FOLFIRINOX. In second line here below, you see that in second line, you basically are pretty much empty-handed with median overall survival times of somewhere between 4 and 6 months, and that's already very, very selective patients because most patients already died. So it stands to reason that with a very good initial activity signal of PELA in combination with atezolizumab with an anti-PD-1, anti-PD-L1, you would explore to see if you can establish a new first-line treatment in metastatic pancreatic cancer. And you could, therefore, consider PELA plus the combo gemcitabine, paclitaxel plus for anti-T cell expansion -- T cell expansion and protection of their effector cell function, atezolizumab, for instance, an anti-PD-L1 or any anti-PD-1 or you work with the PELA plus the FOLFIRINOX combo plus or minus atezo or any other anti-PD-1, PD-L1, so that you would see also the analysis of the contribution of components because how strong was the signal of activity. And this is already my last slide, but look at this. In the GOBLET study that will be discussed in detail after me, there was the observation of a 62% overall response rate, which is somewhere between double and triple what we're used to see with the combo chemotherapies. But even more importantly, I think, is when you look at the spider plot, look at the red lines. So anybody who had a response or very minor response or just stable disease, it is durable. And this is a hallmark of active immunotherapy. Durability is what -- it brings on top of a 62% overall response rate as a signal. And therefore, this needs to be further developed and further explored. And that's where I want to leave it at this point in time. I will be open to any other questions later, but I now first hand over to the next KOL to give his take of the state of the nation. Thank you.

So I'd like to thank Professor Eggermont for his kind of introduction into the role of oncolytic viruses and immunomodulatory potential of this virus. My name is Deva Mahalingam. I'm a GI oncologist with an interest in early drug development in a variety of drugs. And obviously, I have been fortunate to work with the oncolytic viruses since I was finishing my fellowship in 2008 and '09. And so I've had -- it's kind of been working. So it's nice to see this drug come along over the last 15 years. And I'd like to talk to you a little bit about what we have done at least in this space in pancreatic cancer. And I'm hoping I can be with you for the Q&A after this. I'm probably in the middle of the Serengeti at the moment, somewhere in Kenya, but I would have loved to have interacted more in person, but I hope to be present for the Q&A. So my interest in the oncolytic virus stems from the fact that reovirus preferentially infects tumors that are especially KRAS-mutated tumors. And what you can see here on the slide is that the human pancreatic epithelial cells or HPNE cells, when we infect them with KRAS, we see more oncolytic viral replication. And then subsequently, when we did mechanistic work, we found that pelareorep actually induces a strong ER stress in pancreatic cells, as you can see here on the figures with ER dilation on electron microscopy and upregulation of stress markers such as CHOP and PERK. And the disruption of ER homeostasis drives apoptosis supporting the mechanism of action. And obviously, in this slide, we did it in a variety of pancreatic cancer cell lines, and you can see a dose-dependent reduction in cell viability observed, including increase in DNA fragmentation and Cleaved Caspase-3 expression confirming apoptosis. And taken together, these findings suggest that pelareorep induces ER stress, mediated apoptosis and impairs viability of PDAC cells through selective replication to our -- at least when we were working on this in 2008 was felt to be all oncogenic RAS driven. And hence, we focused a lot in pancreatic cancer. And this was the basis of our first study, which was an investigator-initiated but sponsored by Oncolytics known as the REO 017 study, which is a single-arm Phase II study -- single-arm Phase I/II study looking at pelareorep in combination with gemcitabine as first-line trial in advanced pancreatic cancer. And just to remind us that when the study was designed, obviously, gemcitabine as a single agent was the standard of care. The trial enrolled 34 patients, mostly patients with metastatic disease. Most of them had liver metastases. And the cycle was given as a 21-day cycle with gem on day 1 and day 8 and the virus was given on day 1, 2, 8 and 9. The primary endpoint was actually the clinical benefit rate at 12 weeks with the secondary endpoints, including PFS, OS and safety. And the combination was aimed to assess if there's an early sign of efficacy and tolerability in this chemotherapy-naive population. And so REO 017, there were 83% of patients evaluable for evaluation for clinical benefit and most patients had stable disease. And you can see that the spider plot here, you can see disease stabilization with treatment duration extending up to 26.1 months for one of my patients. One of the patients actually did achieve a partial response, while 5 had progressive disease and with a disease control rate of 83% amongst all evaluable patients and 71% in the intention-to-treat population, we felt this disease stabilization and this prolonged stabilization in some patients was, in particular, interesting to kind of look at further studies in this space. One of the things we noted with REO 017 is that when we combined the PELA with gemcitabine, we found a median OS of 10.2 months and with a 1-year survival rate of 45% and a 2-year survival rate of 24%. And this obviously is a significant improvement to historical controls of just gemcitabine alone. These outcomes, obviously, we are seen actually around that time when we were finishing the study with FOLFIRINOX regimens. But obviously, we think that this regimen was better tolerated. The progression-free survival was modest at 3.4 months, but we did see this durable overall survival suggestive of a meaningful long-term disease control in a subset of our patients. Around the time when we were finishing REO 017, there was another study that was done through the NCIC TAP known as the NCI-8601 trial, and this was [indiscernible] study along with Anne Noonan. And obviously, pelareorep was combined with paclitaxel and carboplatin in the first-line treatment. This backbone was acceptable at the time because the only other treatment option at the time was gemcitabine. So they did a randomized trial looking at the PELA in combination with CarboTaxol versus CarboTaxol alone. And in this study, obviously, the primary endpoint was progression-free survival with secondary endpoints being the overall response rate, disease control rate and safety. There were 73 evaluable patients, baseline characteristics were balanced. Of note, 17 patients crossed over from the control to the PELA upon progression. And again, this design was to see if there was some synergy between virus and perhaps a platinum-based chemotherapy. So this slide highlights the key findings from the NCI-8601 study. We see a near fourfold improvement in the 2-year overall survival in the pelareorep arm versus the control, while the median PFS, OS, ORR was similar between the groups, a delayed separation in the survival curves did emerge, suggesting an immune-mediated effect. The 2-year overall survival rate reached 20% in the PELA arm with just 9% in the intention to treat control arm. And if you include the -- exclude the 6% of patients that crossed over, this is certainly meaningful. Obviously, the late survival advantage without the early benefit supports an immunologic strategy and underpins the ongoing trials that we then subsequently did with PELA with checkpoint inhibitors as well. So one of the things we noted is in a lot of the trials that we have done across multiple tumors, including pancreatic, IV administration of pelareorep was detected. We see the virus detected in nearly all tumor biopsies, showing almost 100% presence in pancreatic and breast cancer samples. This demonstrates consistent tumor targeting and replication of the virus. And this IHC image you see here is from a breast cancer patient and highlights the brown staining of the virus in the tumor cells, but not in the surrounding stroma, confirming the selective localization within the malignant tissue. And these data obviously underscore pelareorep's preferential uptake and activity in the tumor microenvironment. And this study is obviously the study which I did, which is the REO 017 study. And one of my patients on the study had this prolonged clinical response. And so we did a biopsy, and I encouraged him if he would be okay, we did a biopsy from the primary pancreatic tumor because we wanted to see whether the virus can actually penetrate the dense fibrosis and get into the tumors itself. And this is what we showed is that like -- so we can see that the virus is replicating in the tumor on the top left. We see it causing caspase activation. But importantly, we see this induction of PD-L1, which offers a compelling rationale of combining pelareorep with immune checkpoint inhibitors in pancreatic cancer that then led to the subsequent trial I'm going to talk to you about next, which is REO-0024, which was then -- the plan of this study was to evaluate the safety and feasibility of pelareorep with pembrolizumab and chemotherapy in the second-line setting, just given the landscape had changed at the time with gem/ABRAXANE and FOLFIRINOX in the frontline setting. So this was a single-arm study that we enrolled 11 patients, median age of 64. A lot of them were female patients. The chemotherapy backbone, we let it be physician's choice because it was a second line between gemcitabine, 5-FU, irinotecan often depending on what they got in the first-line setting. And PELA was given on day 1 and 2 and pembro on day 8 on a 21-day cycle. Obviously, the plan was -- the primary endpoint was safety and looking at the dose-limiting toxicity. But despite the small sample size, we established that this combination was safe, including patients in the second-line setting. And in fact, we enrolled the 11 patients very, very quickly. I think within 6 months. This is some of the efficacy. You can see a lot of patients with one partial response, a few patients with stable disease. And obviously -- so certainly, we know that when added to PELA and pembro, it's safe, and we are starting to see some activity at least in a subgroup of patients that we did here. And as you can see in the waterfall and the swim plots outlined here. Now this led to the next study because one of the questions we had is whether we could drop the chemo and just see the same effects of just the virus and immunotherapy. So 13 patients were enrolled, 12 were evaluable for efficacy. The treatment involved sequentially using PELA on days 1, 2, 3, 8 of cycle 1, followed by pembrolizumab starting on day 8. So we wanted to sequence this stuff as well. None of the patients here that were included had MSI-high. And the study is focused on overall response rate with secondary endpoints, including PFS and OS. And obviously, this study -- as you can see here, we found that PELA and pembrolizumab without chemotherapy in the second-line setting. The objective response rate were modest 8%, but the disease control was about 42% of the patients had a good disease control in this microsatellite stable population. Importantly, immune correlative studies that we had done, which I'm going to show -- which we reported in the manuscript showed that we see T regulatory -- T regs reduction and increase in PD-L1 intensity, especially in responders. And so we believe that this provided this kind of immunomodulatory signature that might provide a combination strategy, especially if we decide to move this combination into frontline therapy with chemotherapy. Of note, this is a -- one of the biomarkers of interest is obviously looking at peripheral T cell clonality. And we actually found that in patients after 1 cycle of treatment, if they had a higher peripheral T cell clonality, they were associated with a better responder and actually improved survival as well. So suggestive that this reactive T cell clonality assay might be a predictive marker for looking at patients that might actually be the survivor and can be incorporated into any trials that we may intend to do in this space as well. So this is my last slide, and I would like to thank you for your attention and hope to -- and this -- I'm glad that this -- a lot of the work here led to the subsequent trial with the GOBLET trial that Dr. Arnold, I'm looking forward to listening to his updated data on the study as well. Thank you very much.

Dear colleagues, my name is Dirk Arnold. I'm Director of the Asklepios Tumorzentrum Hamburg, Germany as well as one of the main investigators of the AIO, which is a collaborative study group. And I do have the privilege to present you some of the most recent GOBLET data with specific attention to pancreatic cancer, which we have presented at the European Society for Medical Oncology Congress in 2023. So what is GOBLET? GOBLET is a basket trial. It's a pan-gastrointestinal cancer trial, and we are now looking specifically to the cohort, which is examining first-line metastatic pancreatic cancer, pelareorep plus gemcitabine plus nab-paclitaxel plus atezolizumab. So it's a full drug combination, 2 immunotherapies, PELA plus atezo and 2 chemotherapies, gem and nab-pac. It's a single-arm cohort. And the primary endpoint of this combination is to assess objective response rate. Secondary endpoints are progression-free survival, overall survival and disease control rate, of course, as well as safety. Let's have a look to the study population. So far, what we can report, they do have a median age of 63 years. They are predominantly male. And notably, all patients had metastatic disease. Nearly 7 out of 10 patients had liver metastases. So what are the efficacy summaries? Before we go more into the granular data, let's have a kind of global overview. The response rate has been 62%, the disease control rate, 85% and 45% of all patients were alive at 1 year. There were no treatment-related safety concerns. And we will also, let's say, lose some thoughts about potential biomarkers, specifically on the expansion of the tumor-infiltrating lymphocyte clones, which seem to correlate with response. But going to more granular data, that's response. So that's objective response and the 12-month survival rate. You can see in this spaghetti plot as well in the waterfall diagram on the right-hand side, the really stimulating data of overall response rate of 62%, disease control rate of 45%. And as I have said before, the 12-year -- 12 months, excuse me, overall survival rate is 45%. So really great signals for efficacy. Well, let's look specifically a bit on patients and responding patients. We haven't seen a clinical pattern of any patient group which benefits more or less. Of course, it's still a limited number of patients, but we can see responses across all metastatic sites, liver, peritoneum, lung and also lymphangitis. So therefore, very difficult to say. Well, obviously, liver metastatic sites seem to be respond as good as others, and that's important. So overall response rate in liver metastases only is 56%. So nearly one by one correlating to the overall response rate. Looking on toxicity and safety signals. Well, predominant toxicity are immunotherapy-related low-range or low-class acute reactions like pyrexia, chills and also some temporary fatigue. There has fortunately been no higher Grade 3/4 treatment emerging adverse events, maybe except neutrophil count decreased, but please have in mind that this is a 4-drug combination of 2 immunotherapies and also a combination chemotherapy. But most common adverse events are flu-like symptoms, GI symptoms and most of them are Grade 1 and 2 or transient. Coming to the biomarker aspect. That's a graph showing on the tumor-infiltrating lymphocyte expansion in patients' blood. What can be seen here is that the pre-existing tumor infiltration expansion in blood increases at cycle 2 to 4 compared to cycle 1 to 2. So obviously, the expansion in blood is a result of the treatment for a certain time. And interestingly, this corresponds with the decrease of tumor size. So it's not a perfect biomarker because we don't see these before we do start to treat patients, but it's maybe correlating to the mechanism of action as we can see that the TIL expansion in blood continues at least through cycle 3 of therapy and the TIL clonal expansion in the blood seems to correlate with a reduction of tumor volume. Thank you for attending.

Hi. Thanks for having me. I'm Sanjay Goel. I'm a medical oncologist at the Rutgers Cancer Institute in New Jersey. Just as a brief introduction, I've been working with pelareorep for probably close to 20 years now, and I'll briefly explain our work in colorectal cancer. So just to give you a brief idea about the burden, it's the second most common cancer among men and women responsible for close to 53,000 deaths in the U.S. alone. And 96% of metastatic colorectal cancers are what are considered microsatellite stable. So I'm not going to go into the details of pelareorep because that's already been discussed by prior presenters. So I'll delve right into the work that we've done. So we started working on this several years ago, and I've highlighted some of the important results. So we looked at several colorectal cancer cell lines that were both KRAS mutant and KRAS wild-type. And we're also fortunate enough to have a cell line that is isogenically similar except for the KRAS gene. And the parent cell line is HCT116 and the cell line that is lacking the KRAS mutant gene is HKE3. So this is essentially KRAS wild-type. So through a series of several experiments, what we discovered was that when we exposed a panel of cell lines, several of them to pelareorep at different concentrations, we find that overall, cells that carry a KRAS-mutated gene essentially are more susceptible to pelareorep than wild-type, as is shown on this graph. To the right is the susceptibility of these isogenic cells to increasing doses of pelareorep as we go from the left to the right. And again, what we can see essentially is that even though mutant and wild-type cells are susceptible to pelareorep, the mutant cells shown in blue have a greater degree of inhibition as compared to the KRAS wild-type. Similarly, when we look at TUNEL staining, which is a marker of apoptosis, looking at the isogenic cells on the left is the KRAS mutant panel that shows a greater degree of apoptosis as compared to the wild-type cell. And finally, mechanistically, we found through western blotting and other experiments that it induces apoptosis in a caspase-dependent manner. We also then look to see whether we can demonstrate synergy with irinotecan, again, with the objective of developing the drug in colorectal cancer. And what we essentially found is that in 12 of the 13 cell lines using a test called the combination index, we found that 12 of the 13 cell lines were -- the combination was synergistic, except in one. This synergy was mediated by p21, as shown in the western blot on the top right. And the synergy was also visible when we tested for apoptosis, showing that the combination has a greater degree of apoptosis, again, both in the wild-type and in the mutant cells. This represents wild-type, this represents mutant. Finally, taking this into xenograft models with the HCT116, this is a classic model. I'm not going to elaborate, but we clearly see a greater degree of tumor control with the combination. So based on this, we hypothesized that possibly pelareorep can sensitize microsatellite-stable colorectal cancer to the effect of immune checkpoint therapy. And essentially, we began a series of experiments. We had 2 syngeneic models, CT26 in which CT26 cells are injected into BALB/c mice. This is KRAS mutant and microsatellite stable. MC38 cells injected into the C57BL/6 mice. This is KRAS wild-type and microsatellite unstable. To summarize what we see is that when we use the combination, it is synergistic as shown in the purple line here. The combination is synergistic in the cell line that is KRAS mutant and microsatellite stable, whereas in the cell line that is microsatellite unstable and KRAS wild-type, the checkpoint inhibitor is highly active as shown by the green line, that the combination really achieves nothing more. So there's no real combination of -- combination activity in the microsatellite unstable cell line. Similarly, we did other experiments showing the synergistic effect in induction of dendritic cells, monocytes and in suppressing macrophages by the combination. Going on further, we show a decrease in Ki-67 an increase in Cleaved Caspase-3 and an increase in TUNEL staining in the combination, but only in the cell line that is microsatellite stable and KRAS mutant as shown on the left panel. Cells that are microsatellite unstable and KRAS wild-type, we do not see the synergistic effect. So to conclude, I would say that pelareorep is active in KRAS mutant and wild-type cells, but more so in mutant. It is synergistic with irinotecan and it's mediated by p21, and pelareorep sensitizes microsatellite stable cancers to the effect of checkpoint therapy and is synergistic with immune therapy. Going on to a clinical trial in brief. This was a combination of FOLFIRI and bevacizumab in oxaliplatin refractory KRAS-mutant colorectal cancer. We performed a Phase I study looking at escalating doses. And even though initially, we saw 2 dose-limiting toxicities in patients who had prior exposure to FOLFIRI, we amended the protocol to only include FOLFIRI-naive patients. And at the highest dose of irinotecan, which is 180 and the highest dose of pelareorep, which is 3x 10 to the 10 out of the 6 patients, no DLTs are observed. Interestingly, in these patients, the overall survival was 25 months, as historically, one would expect only 11 months, and the progression-free survival was 15 months. Historically, we would only expect 6 months. In terms of some correlative studies, we saw that there's a clear induction of dendritic cells as shown by an uptick in CD123 at 2 days, followed 5 days later, but an increase in what is called CD8/CD70. So these are activated CD80 lymphocytes or really the active T cells. This is just a comparison of the 3 randomized Phase III trials of FOLFIRI plus an anti-VEGF agent, comparing them to the data that we observed. As I said, we've seen 25-month survival. All the others saw between 11 and 13. Our PFS was 15 months. At best, what was seen before in this Phase III trial was 7 months. And our response rate was 50%, where previously the best response rate was 20%. So even though our data is limited, I think there's clear evidence that the combination of FOLFIRI, bevacizumab and pelareorep is highly active. So I'm going to skip through this. So essentially, what I hope to do is launch a randomized Phase II trial of metastatic microsatellite stable KRAS-mutant metastatic colorectal cancer to give them -- to randomize them into either FOLFIRI, bev and pelareorep, which is really the clinical trial that I just presented in terms of the Phase I and randomize them to either the combination of standard of care versus the combination plus a checkpoint inhibitor with the primary endpoint being response rate. And essentially, we would need 28 patients in each arm, and we expect a response rate of about -- I have it here. The null hypothesis is a response rate of 15% in the control arm and 39% in the arm with checkpoint inhibition. And I'll end there.

Thank you, everyone. Thank you for those presentations. A lot of really interesting information, obviously. And so I think right now, we want to jump right into the Q&A. Some people call it a fireside chat. I sort of dislike that term, but mostly because there's no fireplace around. But I want to get into questions. I think we have the First one coming in from an analyst, Michael Freeman from Raymond James. Sort of a broad-based question for Dr. Arnold here. Sort of based on the data we have to date with PELA in the first line of PDAC, does it follow that PELA should be explored in a registration-enabled study in the first line?

Well, yes, good question. I think that's the key question. Well, we have seen this really for us also promising and unexpected efficacy. And given the unmet need in pancreatic cancer and the need for improvement, which Professor Eggermont has nicely introduced in his talk, I think we all would really die for having a kind of registration trial, which is aiming to improve overall survival and also symptom control for patients with pancreatic cancer. And I think that's one of the beauties of the GOBLET trial. We have seen this great 1-year overall survival rate, which is really stimulating. But I think the most stimulating thing is this shrinkage of tumors, let's say, the lowering of the tumor burden. And this is in most patients associated with relief from tumor-related symptoms. So therefore, we would, in any case, like to offer patients with metastatic pancreatic cancer a great and hopefully very efficacious treatment. So therefore, I would give a thumbs up for this thought. And I would dare to do this and to see this.

And one quick follow-up, Dr. Arnold, and then others can jump in too. And Tom -- Dr. Heineman, Tom, interested in your take on this. But what would be the sort of just a couple of key points of that study, primary endpoint, maybe a hypothesis on the size of the study? What are your takes on that? Dr. Arnold?

Well, traditionally, when we see registrational trials in first-line pancreatic cancer, it's always an overall survival endpoint. And I think that's really a challenging thing that we have to show in a randomized trial that this combination can improve or will improve the overall survival likelihood. So therefore, we talk about, let's say, many hundreds of patients. I don't see, let's say, thousands of patients, but it's clearly more than 200, 300 patients, which we have to see -- which we have to invest. And also, we have to see that the treatment landscape in pancreatic cancer will be evolving over the next few years. We will have all of these RAS inhibitors coming up also, let's say, coming into the market of first-line treatment. So therefore, we have to, let's say, discuss the environment where we evaluate this. But it's a randomized Phase III trial. We have to have a randomized Phase III trial in pancreatic cancers due to the methodology, the registrational agents are -- or agencies are asking for, and it has to be an overall survival endpoint.

Great. Yes. And we'll get to the treatment landscape shortly. But Tom, on this point, we have this GOBLET data with the checkpoint inhibitor which would sort of bring into this component -- contribution of components thesis here. But our data, we have strong data with just a chemo plus PELA backbone. So quick thoughts on 2-arm versus 3-arm study, Dr. Heineman?

Yes. So I think that the data we have are actually promising -- show promising efficacy for PELA either both with and without a checkpoint inhibitor backbone -- I mean component, right? And so I think you could make a very good case for investigating going either pathway. Of course, if you go without a checkpoint inhibitor, it makes a simpler and probably smaller study. If you go with a checkpoint inhibitor, you would have to add additional patients to sort out the actual contribution of the checkpoint inhibitor component. So I think it's actually a good problem to have. There are a couple of different ways we could go. And I think both of them are valid, and we'll just have to see where we'll have to agree on the best design with our experts and also internally, but we have 2 good options, I think.

Yes. And Dr. Mahalingam, it looks like you may have technical difficulties. But if you can hear this question, I wanted to throw you a quick follow-up. What is the -- we have 1,100 patients worth of safety data. What does that tell you about PELA's safety profile?

I think he dropped off...

Okay. We lost him again. How about Dr. Goel. 1,100 patients' worth of safety data in PELA across multiple tumor types. And what takeaways can you make on the safety data?

Yes. I mean I think I've had experience with this starting from the single-agent Phase I and to several other trials. And of course, looking at all the data, it really seems that the drug, if you call it a drug, is really overall quite safe. I would not attribute any major additive toxicities to the drug other than what you'd expect when you get a viral infection, like just some fever, mild diarrhea, et cetera. So I think overall, it's tolerated very well.

Yes. And how often do you have a late-stage biotech with 1,000 patients' worth of data before a drug approval? How often do you see that?