Replimune Group, Inc. (REPL) Earnings Call Transcript & Summary

Good morning, and thank you for joining our investor conference call. We're excited to have this opportunity to share with everyone our first data set from a study being run with the registrational intent. Before we begin, I need to highlight our forward-looking statement. On Slide 3, I'm Philip Astley-Sparke, CEO of Replimune. And I have with me today from management Robert Coffin, our President and Chief R&D Officer; and Sushil Patel, our Chief Commercial Officer. We're also very pleased to have with us today several leading physicians from sites, all with considerable experience of using our products. Dr. Mark Middleton, PI on the IGNYTE study, for which we are reporting data today. Mark is Professor of Experimental Cancer Medicine and Consultant Medical Oncologist at the Oxford Cancer Centre and Head of the Department of Oncology at the University of Oxford. Dr. Mike Wong, Professor in the Department of Melanoma Medical Oncology at MD Anderson. And Dr. Kevin Harrington, Professor in Biological Cancer Therapies at the Institute of Cancer Research, London, and consultant clinical oncologist at the Royal Marsden. Slide 4 shows the agenda for the call, which will start with a brief further introduction from me, followed by 3 content sections. The first section covers the data from the first 75 patients of our 125-patient study in anti-PD-1-failed melanoma being run with registrational intent. The data speaks to the emergence of a potential new treatment option in this unmet need setting with a 36% overall response rate, a 20% complete response rate, strong durability to date and strong data in all stages of disease, including late stage 4. The second section will speak to the commercial opportunity as we seek to establish a broad skin cancer franchise with RP1 before providing an update in section 3 on RP2/3. Here, we will disclose which of RP2 or 3 we plan to bring forward in relation to our previously disclosed Phase II program in CRC, HCC and head and neck cancer. We are also excited to announce today a cost-sharing collaboration with Hoffmann-La Roche and the potential fast to market indications of third-line CRC and second-line HCC. In addition, as proxies for success in hard-to-treat immune insensitive tumor types, we'll be presenting some additional data in uveal melanoma and sarcomas. Following the presentation, we'll be holding a live Q&A. To ask a question, you'll note there's a question submission bar on your screen that will be open throughout the presentation this morning and during the Q&A session. We will collect those questions throughout the morning and take as many as time allows. All questions will be held until the Q&A portion of the event. On Slide 6, as a reminder, we are the leader in tumor directed oncolytic immunotherapy with 3 wholly owned programs. As aforementioned, we plan to establish a broad-based RP1 skin cancer franchise and today have taken a giant step towards realizing that ambition with strong melanoma data in patients with limited other options. We now look forward to revealing data in our randomized registrational study in frontline cutaneous squamous cell carcinoma in the first half of '23, having generated a high rate of complete response in a prior study. With a broad-based mid-stage development plan now in place for RP2/3 beyond skin cancers, the company has the potential to generate substantial revenues beginning in 2025. We have made solid progress, putting in place the building blocks to build an entity capable of transforming the immuno-oncology landscape. Our own manufacturing facility is fully operational. GMP batches have been released to the clinic, and we've shown comparability to the contract material we've used in our studies to date. And our BLA consistency lot campaign is now in full swing. In addition, our Chief Commercial Officer and his team are continuing the planning process to ensure, if approved, our products are widely adopted in the marketplace. Finally, we have a strong balance sheet to execute on our vision with runway into 2025. Next slide. As a reminder of our technology in MOA, tumor directed oncolytic immunotherapy is the use of viruses that when injected into tumors partially or completely destroy them through virus replication, bursting the tumor cells open and exposing all the release cancer antigens to the immune system in an environment of necrotic cell death. This leads the activation of a patient-specific systemic immune response and the destruction of uninjected deposits. On Slide 8, our platform is based on the herpes simplex virus, which we believe is an optimal viral species for oncolytic use. They're being both highly lytic and inflammatory and having the ability to carry multiple immune-stimulating proteins into the tumor microenvironment to further amplify the immune response generated. Our strain of HSVs have been deliberately selected for its lytic properties in human tumor cells. And from all of our products as our base platform, we express a fusogenic protein that greatly increases direct tumor killing, immunogenic cell death and systemic immune activation. These design features ensure maximum presentation of antigen on the MHC of antigen-presenting cells. We then express various proteins from the virus intended to maximize co-stimulatory signals of the APC/T cell interface to ensure optimal T cell priming. Our lead product RP1 additionally expresses GM-CSF. RP2 additionally expresses an anti-CTLA-4 antibody to stop the negative feedback loop of the CD28-CTLA-4 axis. And RP3, too, further immune co-stimulatory pathway activation proteins targeting CD40 and 4-1BB, which also leads to downstream inflammatory cytokine release. Our confidence in our ability to establish a broad skin cancer franchise with RP1 is underpinned by broad activity across a variety of skin cancers as monotherapy and in combination with checkpoint drugs, including in the checkpoint refractory setting, where clear systemic long-term responses have been generated now in a substantial number of patients. Moving to Slide 9. We are continuing, therefore, to make great progress towards our ambition of making our products a cornerstone of immuno-oncology treatment regimens as the most practical and effective way to initiate a systemic antitumor immune response. We're operating in only a small number of significant white spaces left in immuno-oncology, where we have the potential to have transformative benefit: firstly, checkpoint failed disease; and secondly, patients with liver metastases. On Slide 10, the benefit now shown in anti-PD-1-failed skin cancers from multiple cohorts is now clear and obvious, and we've also seen anecdotes in multiple other non-skin cancers that we can benefit checkpoint-failed patients. The data set we are presenting today from 75 patients is consistent with all our previous data sets in anti-PD-1-failed disease with a 36% overall response rate and a 20% complete response rate and points to the emergence of a new treatment option in this setting. On Slide 11, we also look forward to seeing if we can broadly benefit patients with liver metastases, where I-O drugs have [ proved to be ] ineffective and prognoses are poor. We have seen multiple anecdotes of being able to benefit patients with liver tumors across a number of cancer types and look forward to progressing studies in HCC and CRC. Today, we announced we'll be collaborating with Roche in these indications, validating our approach. I will now hand on to Dr. Middleton, the PI on the IGNYTE study to summarize today's data set in anti-PD-1-failed melanoma.

Thanks very much, Philip. So next slide, please. So I'm going to talk about second-line melanoma. And the landscape at the moment here is that there are no good options for our patients who progressed on anti-PD-1 therapy. You can have anti-CTLA-4 treatment, either as a single agent or in combination with PD-1. But the toxicity here is quite significant. And the benefits is, we've seen, somewhere between 1 in 3 to 1 in 6 patients depending upon circumstances. Although we're seeing a change in the landscape with the use of anti-LAG3 agents, this is very much in the first line as a mitigation of toxicity for combination immunotherapy, and there's no meaningful activity that I'm aware of in anti-PD-1-failed melanoma patients. We have the option in about 40% of our patient population of charting the MAP kinase pathway, but although that's fairly successful, its transient responses last to the order of about a year before resistance develops. So it's not a long-term solution for patients in the way that I-O promises and has delivered, albeit for a minority. And there are other options available that are experimental, in particular cellular therapies such as TILs. The response rates here are about 1 in 3 and showing some promise, but it's logistically enormously complex, not suitable for all of our patients and comes again at the cost of considerable toxicity. So overall, for the second-line melanoma population, leaving aside that kinase inhibition, about the best one can get benefit in about 1 in 3 of the population, but at considerable expense, both in terms of logistics, toxicity and finance. So moving on to the next slide. What I'm going to talk about here are the data from the IGNYTE registration cohort of anti-PD-1-failed melanoma and then link that with prior reported PD-1-failed melanoma patients from our Phase II cohort in the earlier part of the study. Median follow-up is now just short of 10 months, and the data I'm going to present you includes a variety of patients, including those with moderate and high tumor burden and the majority of patients have not responded to prior anti-PD-1 therapy, so had, had stable disease for refractory in the first instance. We're going to show you some responses across all subgroups. And I think it's interesting to note that although we see the expected difference in response rate according to the extent of disease, we still see responses across all subgroups of melanoma. And the vast majority of those are ongoing, so in terms of the benefit [ that's powered ] to add from the snapshot today. We also see responses in uninjected lesions, the so-called abscopal effect, which includes patients with visceral disease. So we're going to see examples of responses in solid organs like liver in patients who've been injected with superficial lesions, for example. The other thing to note is that this combination is pretty well tolerated. We had only Grade 1 and 2 on-target side effects attributable to RP1 and the expected toxicities of the nivolumab with the relatively low rate of toxicity that we understand from that agent. Survival data with the progression-free overall is not yet mature, but we're seeing some interesting trends, and I'll cover that right at the end of my section of the talk. Moving on to the next slide. This summarizes the way which we're treating the patients in the cohort with -- starting with the RP1, increasing the dose and adding in the nivolumab at the second cycle 2 weeks later and continuing with 8 injections of RP1 and then the nivolumab going for up to 2 years with the possibility of reintroducing the RP1 during that time, if necessary. I draw your attention bottom right to the criteria for having failed on checkpoint inhibitor treatment. These are quite strict criteria requiring at least 8 weeks of prior treatment and confirmed progression whilst on anti-PD-1-based therapy. Notably, we do include patients who have progressed whilst on adjuvant therapy, requiring confirmation of that progression before they could be enrolled in the study. Next slide. So these are the demographics. On the left-hand column, you can see the original 16 patients from the early part of the study. In the middle, the 75 we're reporting on now. And on the right, the combined cohort. Would draw your attention to here is -- this is a relatively representative population of PD-1-failed melanoma. This is a significant proportion, about 1/3 of patients who have come having only had PD-1 as prior adjuvant therapy rather than for metastatic disease. And I'll come back to that later, but we see a good bunch of the patients, half in total, who've got M1b and c disease, so proper metastatic disease. And in the current cohort, a slight increase in the proportion of patients who've got an elevated LDH, which as you know is an adverse prognostic indicator in this setting. Next slide. So the toxicities, and I'd refer you to the Grade 3, 4 and 5 columns in the middle of this table are really quite mild. So very small numbers in those columns. Some of those, but the minority attributable to the RP1 in the green box there. And the majority of the higher-grade toxicities, those that we expect to see with nivolumab, continued over a 2-year period. So overall, the message here is that there's very little in the way of additive toxicity compared with nivolumab alone, and it's very well tolerated. Of note, this is all the skin cancer patients being treated with RP1 and nivolumab across the trial program. So this is a higher number than the 91 that we're talking about in the PD-1-failed melanoma cohort. But we don't expect toxicity to differ between other skin cancer patients and those in that cohort. Next slide. So here's a snapshot of the data looking at the response rate, split on the left-hand side between the original cohort and today's 75-patient snapshot. Of note, within that group, over 1/3 of patients responding, 36% objective response rate and over half of them showing evidence of disease control. I think a key point to note here, as we scan across to the right on the table, is that responses are seen across all settings, whether that's patients who've had PD-1 in the adjuvant setting as their prior therapy or those who have had PD-1 as part of treatment for metastatic disease. It's also the case for those patients who've had CTLA-4 in combination with PD-1. And we see good responses in patients with late Stage 3 and early Stage 4 disease, but also high response rates for patients with Stage 4b and 4c disease, which often has been refractory to experimental therapies. Next slide. Showing these data in a different way. This is the waterfall plot. And I think the key point to take away here is the depth of response that we see, with all those bars on the right-hand side going down to minus 100, indicating complete response in target lesions. Over 50% of patients have had a reduction in their target tumor diameters. And as you can see if you dig down into the detail, I'd draw your attention to the green and the orange bars, complete response is being observed in those patients with M1b and c disease. Next slide. Looking at these data in a slightly different way with the spider plot. The key takeaway from this is that responses have the potential to deepen over time and are proving extremely durable with many of these curves out beyond the 2 years of treatment. We're approaching that. Next slide. And looking at this in one final way, with a swimmer plot, what we're showing here is the durability of those responses. If you look at the very top, 720 days is 2 years of treatment, give or take, 730. And you can see responses going out beyond the end of treatment. A lot of patients continuing on treatment at earlier time points, suggesting that we're seeing very durable responses. And on the next slide, we can see that these obtain where the patients have had relatively short and ineffective prior PD-1 therapy or can also be obtained in those who've been on PD-1 for quite some time before failing that and moving on to the study. Again, 85% of our responses are ongoing. So the duration of response is likely to increase over time or the median duration of response. The majority of those responders are now already out over a year, despite this being early snapshot of the data. Next slide. We see responses in uninjected lesions as well as injected lesions. So here, we've got the numerical consideration in blue of the lesions that have been injected. Many of the patients, it's only a single lesion. But in some, particularly those with superficial disease, we can get multiple regions, with 13 being the highest. That being said, we also see responses in uninjected lesions or stable disease in lesions that haven't been injected, as indicated by the orange and green bars. And the majority of responding patients have lesions that we haven't been able to inject. So we can get responses systemically by treating a minority or a proportion of the accessible disease. Next slide. And I said earlier on that we would talk a little bit about the survival data, although these aren't mature. But I think what one can take away from this, despite the censoring and the relatively early cut of the data, is strong evidence for plateauing. And those patients who get through the first couple of months of treatment with that progression have a high chance of remaining on treatment. And that's borne out with impressive PFS interim statistics and also the overall survival that we see there, although this is immature. Next slide. So I want to finish by talking a little bit about some of the cases that we've seen in the trial to demonstrate some of the important points that come from these data. So this is a patient from Utah who's been injected in a single lesion on the forehead, as shown here. And throughout the slides that follow, injected lesions will be ringed in red and uninjected lesions will be ringed in yellow. So you can see here that over the course of 9 months, there's almost complete resolution of the injected lesion. Next slide. But more impressively and importantly, when we look at visceral disease that's uninjected, surrounded by the yellow ring, we can see impressive responses on the left in liver and, on the far right, in mediastinal lymphadenopathy. These patients have been exposed to both pembrolizumab and to nivolumab for their stage M1c disease. And we also see responses in lung, as shown here, which are deepening over time. Next slide. And in a second patient now, treated in Leeds here in the U.K., a patient who had both combination and single-agent PD-1 exposure in the past with M1b disease, my colleague, Adel Samson in Leeds injected the subcutaneous nodule shown here on the left-hand side at baseline. But we were able to obtain responses in both the injected lesions and uninjected lesions subcutaneously. And on the next slide, in the lung too with a pretty sizable region at baseline, showing considerable diminution in size on a year of treatment, referring you over the pictures on the far right. Next slide. A further patient who had only had pembrolizumab as prior adjuvant treatment, again, this is a nodal lesion that's been injected, but we see responses both in that lesion and also uninjected lymph nodes at multiple locations. Next slide. And showing that in considerably more detail and looking at the responses going out to a year in the evidence that we can present today. Next slide. Then finally, one of my own patients, a patient who progressed on anti-PD-1 treatment for their stage M1c disease, and it also had prior BRAF and MEK inhibition. And he had spinal metastatic disease, which you can see bottom right, as well as a subcutaneous lesion in the groin, which we injected, which is there on the top left. And although that was the only lesion that we injected and ultimately we got a complete response in that, which evolved relatively slowly over the course of the year, we saw a resolution of his lung disease, shown bottom left, improvement in his spinal disease. It's perhaps best appreciated on the central views on the bottom where you can see it disappear, although the kinking spine didn't recover. And I saw him actually only last week where he remains disease-free. Next slide. So in summary, the combination of RP1 with nivolumab continues to be pretty well tolerated with transient Grade 1-2 side effects attributable to the RP1 and in no way adding to the side effect profile that we expect with nivolumab. Over 1/3 of patients in our cohort have shown objective responses to treatment. And 1 in 5 of the patients in the cohort, and that's in the whole cohort, not in the responders, so a pretty impressive statistic, that have shown complete responses to treatment. 5 out of 6 responses are ongoing and durable to date, and most of these responses have come in patients who did not respond to prior anti-PD-1 therapy as opposed to those who did respond and then lost that response. So the combination of RP1 and nivolumab has pretty meaningful clinical activity across a range of situations in which patients who have failed prior anti-PD-1 therapy, whether that's in the adjuvant setting, whether that is multiple lines of treatment for metastatic disease, either single agent or in combination with CTLA-4 and includes patients with M1b and M1c diseases. We've seen responses in patients in both injected and uninjected lesions, including, as I showed in the last few cases, in visceral disease. And overall, the progression free and overall survival look promising, although currently relatively immature. Next slide. So that concludes my portion of the talk. We're saving questions to the end. So I'm going to hand over to my colleague, Sush, who's going to talk about the commercial opportunity.

Thank you, Dr. Middleton, for that comprehensive overview of the IGNYTE data. It's certainly exciting to see the snapshot data is very much in line with the target product profile we've tested with customers and consistent with our ambition to provide transformative benefit for skin cancer patients. So over the next few minutes, I'm going to share the market opportunity for RP1 in skin cancer and, importantly, explain why we believe we are able to realize that opportunity for a large proportion of patients. But let me just start by putting the data Mark just reviewed into a commercial perspective. So firstly, the data demonstrates, we are able to help a range of anti-PD-1 failed patients that occur in routine clinical practice. And I think importantly, one involving area and growing area and usage of anti-PD-1 is in the neoadjuvant and adjuvant setting. And despite this approach curing many patients, unfortunately, about 1/3 of patients will relapse within a year. And there are a portion of those patients that really blow through treatment quite quickly, often viewed as sort of those patients who progress in less than 6 months, and there really is a need for better options for those poor prognostic patients. And as seen by the data today, we believe RP1 in combination with nivolumab provides a compelling option there as well. So I think another important aspect as we think about broad utilization of the treatment regimen is not only whether it works in a broad range of patients, but one that's well tolerated. Many currently used treatments, including combination regimens and emerging approaches such as TILs had a significant patient toxicity burden. And so that's clearly an area of opportunity for us. Given the RP1 tolerability profile, we believe this is an area where we can grow the treatment market as many patients unfortunately forgo treatment due to toxicity today. Next slide, please. So starting in the middle of this graphic, given the anti-PD-1-failed melanoma data generated today, we believe we can help a significant portion of the patient population, including those patients who have failed prior adjuvant treatment and second-line plus patients regardless of their BRAF mutation status. Or said another way, we have the opportunity to help patients no matter what part of the treatment journey they are on. And as mentioned, given our tolerability profile, we can help grow that treated market to approximately 13,000 patients. The IGNYTE data also builds on the strong signal we have seen in cutaneous squamous cell carcinoma, where RP1 also has the opportunity to help many more patients, including those where existing checkpoint inhibitors are not indicated such as immunocompromised patients or those who have failed checkpoint inhibitors. So given the high complete response rates we've seen for RP1 in CSCC, we have the potential to change the treatment mindset and expand the treated population of advanced CSCC in terms of those who are currently deemed ineligible for curative surgery or radiation. And again, this could amount to another 10,000 additional patients. Finally, the higher ORR and particularly the higher CRs we've seen in both cutaneous squamous cell carcinoma and checkpoint-failed melanoma means that RP1 presents a logical partner for new and existing treatments used in the early disease setting. There's a lot of drug development going on in this space, and there's also a lot of patients. Next slide, please. Skin cancers represent the ideal setting for tumor directed treatment with RP1, as a high proportion of patients have a superficial lesion that can be easily injected without image guidance. In CSCC, this is around 90% of patients, as the disease typically presents local regionally. But we also believe the unmet need and activity seen in visceral disease with RP1 in IGNYTE data, as shown by some of the patient cases that Dr. Middleton shared with you, can really help us achieve strong penetration in these melanoma patients with simple and widely available ultrasound guidance, where we can then tackle approximately 70% of the patients and provide that opportunity and benefit to a much broader range of patients. Next slide, please. So our mission is to ensure as many patients as possible can benefit from RP1-based treatment. And to realize that ambition, we aim to launch broadly in the community across practice settings. We have done a significant amount of market research to understand what it will take to achieve this goal and believe strong data will result in the strong adoption and are working actively to ensure that there is customer confidence in RP1 and that the administration can be done routinely and result in a positive experience. There are 3 key areas as we think about facilitating broad adoption. Firstly, we are planning for a product that can be stored at refrigeration for an extended period rather than as opposed to minus 70 or minus 80 degrees in cold storage. We also want to ensure that we identify and train injectors at all our key target accounts. And then finally, we want to make sure that we're educating on how to safely and efficiently handle RP1 to make things as easy as possible for patients and the health care team. Next slide, please. As discussed, our critical success factors are about ensuring customers have confidence and a positive experience with RP1. And our goal is broad community launch and adoption. And there are many reasons we believe we are set up for success with this objective. It starts with transformative data, in particularly the high PRs and CRs that are durable, which is clinically meaningful for both our initial skin cancer indications. We also have strong U.S. clinical sites and patient enrollment in our skin studies. And there is existing customer and key opinion leader experience to build on and leverage as we start our launch activities. Finally, as we consider the important aspect of practice economics, especially for community MDs in the U.S., it's important to remember that we are adding RP1 on to existing checkpoint inhibitors versus replacing them. And this aligns with the current buy-and-bill reimbursement model. Next slide, please. Finally, what's necessary for broad adoption is the availability of sufficient and reliable supply of RP1. And here, we're making really good progress on manufacturing of commercial product, including tech transfer, comparability analysis and initiating BLA consistency batches. Our state-of-the-art facility will produce sufficient RP1 to meet global demand. And it should be noted that RP1 comes with very attractive cost of goods. And now I will move over -- we'll now move to the next part of the presentation, which will focus on updates to RP2 and 3. And with that, it's my great pleasure to introduce Professor Kevin Harrington, who as a reminder is at the Institute of Cancer Research in the U.K. and a renowned expert in head and neck cancers, skin cancers and novel therapeutics. And I think he is coming to us live from ESMO I-O. So over to you, Kevin.

Sush, thanks very much indeed and thank you for this opportunity to share some thoughts around the development of the RP2 and RP3 platforms. And indeed, you're correct, I am joining you from ESMO I-O. So I hope if any extraneous noises occur, you'll understand they aren't entirely under my control here at this meeting. So if I could have the next slide, please. So I'm going to talk to you, first of all, around the area of RP2 in the treatment of patients with uveal melanoma, and I'm going to give you some context to that, if I may. So you may be aware that ocular or uveal melanoma is a rare cancer occurring in about 1,000 individuals per year in the United States. It arises within melanocytes that exist within the normal eye structures in a number of places, including in the iris, the ciliary body and also in the choroid as sites that can occur for this condition. Now the nature of these diseases is a poor prognosis of tumor with a median overall survival of about 12 months. And again, many of you will be aware that uveal melanoma behaves entirely differently from cutaneous melanomas. It is a highly metastatic disease with a dominant metastatic burden occurring in the liver in up to 90% of cases. And once this occurs, historically, this has been associated with a very short prognosis, with only 10% of patients surviving a year or more with that occurrence. Historically, even in the era of the checkpoint inhibitors, this has proven to be a very difficult-to-treat disease with limited activity of checkpoint inhibitors in a number of studies. Again, you may be aware that there has been a first approved agent in patients with uveal melanoma, which is tebentafusp or KIMMTRAK. Now that agent is specific for those with HLA-A-02 disease, which is about 50% of the total population, especially in a Caucasian population. So there exists a continuing unmet need in uveal melanoma patients. And there is a requirement for improved efficacy and tolerability. The treatments that I've just discussed very briefly are associated with significant toxicity. And also, we have the unmet need for those patients who are HLA-A-02 negative. We also have to consider the treatment of patients who have had KIMMTRAK or anti-PD-1 therapy and have either responded and then progressed or have failed to respond primarily. Next slide, please. So here, I show you the initial experience that we have of patients treated with RP2 for patients with uveal melanoma. And I show you here in this rather busy slide, and I'd like you to focus in the first instance, if you will, on those patients highlighted with the green color. So this is the first 14 patients for whom the outcome is known in terms of their response to RP2 therapy, and all of these patients are in that anti-PD-1-failed subgroup of patients. 4 of the 14 patients have developed responses. Many of those responses, and I'll show you data from these, are deep responses and highly durable responses, including clinical complete responses. You will see highlighted at the bottom of the slide details of 3 other patients who are ongoing, all of whom have liver, one with liver and lung disease for whom we do not yet have the full characterization of their response, 1 with stable disease, 2 patients in whom the response is not yet evaluable. But again, notice all of these patients have received an anti-PD-1 therapy, and many of these patients have also received an anti-CTLA-4 drug. So this is a group of patients for whom there is a significant unmet need. Next slide, please. So here, I show you some data of a patient who had previously been treated with both nivolumab and ipilimumab, patient receiving RP2 as a monotherapy agent, not in combination with an immune checkpoint agent. And you can see, again, adopting the schema that Professor Middleton showed you, injected lesions highlighted in red, uninjected lesions highlighted in yellow, you can see that this patient with very heavy burden of liver disease. And on these slices from the scan, I show you some of the lesions, by no means all of the disease, that this patient experienced. And as you can see, over time from screening through 3, then 6 and then 9 months, the evolution of a very solid partial remission. Regrettably, the patient progressing at 15 months. But nonetheless, evidence of response both in injected and in uninjected so-called abscopal or anamnestic lesions. Next slide, please. This is a patient who I treated within my own practice. This is a patient who had received prior nivolumab therapy. You can see that this patient had undergone a nucleation of her left eye. The patient regrettably had developed recurrence in the orbit, an extremely difficult-to-treat area that had been attempted to be addressed with surgery. We had also, in the past, had seen this lady for radiation therapy. And she came to us really with a very difficult-to-treat problem, and she was enrolled in the study where she was treated with RP2 plus nivolumab. And again, you can see from the screening scan at the top, the injected lesion through to the scans that evolved over time to nearly 2 years a clinical and radiological complete remission. Now this patient had disease in the neck, which was uninjected and showed a response and also had bone metastatic disease uninjected. And you can see that over that nearly 2-year period, resolution of soft tissue, but really remarkably also healing of the bone with recalcification of the bone. This patient had a dramatic symptomatic benefit and, on a PET-CT scan, has an ongoing metabolic complete remission at nearly 2 years. Next slide, please. Here, I show you another patient, again, from a colleague, and this is from the team in Liverpool. Here is a patient again with liver metastatic disease, prior treatment both with anti-CTLA-4 and anti-PD-1 therapy. You can see on this MRI scan initially at screening in December of 2021, so a year ago, and then sorry -- could we go to the next slide, please? Thank you. You can see the initial screening scan from about a year ago and then the follow-up scans about 9 months later. And you can see, in the various sites across the liver, the evolution of an impressive partial remission of this disease, including responses within uninjected lesions. Next slide, please. I'd now like to switch gear and give you a flavor of some of the emerging data that we're seeing in patients with soft tissue sarcoma disease. Now this is an area of considerable unmet need, over 13,000 cases per year in the United States this year. The standard of care following surgery includes radiation, often followed by chemotherapy. Anti-PD-1 therapies are used in some types of sarcoma, but not many, and these are in the relatively responsive subtypes. You may be aware that in this area of high unmet need, the FDA has approved a drug, trabectedin, YONDELIS, for unresectable or metastatic leiomyosarcoma and liposarcoma having progressed on anthracycline-based chemotherapy. And that approval was based upon an overall response rate of 7% versus 6% and a progression-free survival of 4.2 versus 1.5 months. So I think you may conclude that, that's a relatively low bar. We also can consider that for many or most subtypes of sarcomas, overall response rates in the range of 25%, or indeed often lower, would be considered as promising, especially in the second-line setting or later. And there are several subtypes of this disease for which there is no FDA approval of drugs. And the NCCN lists a number of agents, not necessarily on evidence base, but that can be used. Many of the subtypes are documented as being resistant to anti-PD-1 therapies. And indeed, anti-PD-1 therapy in a single agent or combination remain unapproved for many of these diseases, hence a significant unmet need. So we could conclude that there are tumor types here where single-arm data based on this unmet need with a strong overall response rate and durability of response might yet be suitable for approval. And indeed, that is an opportunistic, data-driven development of RP3 that we would consider. Next slide, please. So here, I'm going to show you data from the first 5 patients who have been treated with RP3 in combination with nivolumab. You will see they represent a range of these histotypes that we see epithelioid sarcoma, leiomyosarcoma, myxofibrosarcoma, osteo and chondrosarcomas. All of these patients had failed prior therapies, standard-of-care surgery, chemotherapy and other therapies. And so far, for the first 3 evaluable patients, all of them have shown a response to treatment. On the right-hand side of this slide, I showed you briefly the appearances of those lesions. But now I go on to a case, patient treated under my care, a patient with metastatic epithelioid sarcoma originally arising in the perineum. The patient relapsed pleural disease and chest wall disease. So this is a view from the lateral side of this man's right chest wall and you can see horrible disease, patient with significant pain, recurrent pleural effusions, received treatment with RP3 in addition to immune checkpoint inhibition with nivolumab, and you can see for yourselves a dramatic response. The patient did not require analgesia and did not require further pleural effusion drainage, revolutionizing this man's life. I draw your attention to the fact that in an ESMO publication, 80 patients treated with rare types of sarcoma, there was only a 15% partial response rate, no CRs in that study with single-agent pembrolizumab. Next slide, please. Here, I show you another patient. This is a man with multiple recurrent leiomyosarcoma who had undergone at least 12 surgical procedures. You can see the battle scars for that. In -- this is a view of this man's back of his leg at the level of the thigh. This man was approaching needing an above-knee amputation for this disease when the surgeons referred him to our care, and he received RP3 therapy in combination with immune checkpoint blockade. And you can see for yourselves, over a relatively rapid period of time, a dramatic response of his disease such that this man that has no evaluable disease. Next slide, please. And here, I show you the final case, which is a patient with a mix of fibrosarcoma. You can see this glows very brightly on the MRI scan. And again, you can see in multiple injected lesions and, on the right-hand side, you see the clinical evolution of a dramatic clinical response. This patient with disease in the forearm, in his dominant arm, he's a right-handed man with the right-arm sarcoma, was facing an above-elbow amputation, a catastrophic surgical procedure. He has enjoyed an extremely good clinical response, is left asymptomatic with excellent healing of these lesions and he's on the way, we hope, to a solid, indeed complete remission. Next slide, please. So with this, I turn the presentation back over to Dr. Rob Coffin, and I thank you for your attention.

Thank you very much, indeed, Kevin. So I'm now going to summarize our current plans for the Phase II development with RP2 and RP3 that's intended to initiate next year. So this doesn't currently include any further expansion in uveal melanoma or in sarcoma, as Kevin has just described, despite these clearly being very exciting opportunities where we have seen in a relatively small number of patients admittedly, clear and meaningful clinical activity. So that's for the future. So for today, I'm going to focus on the Phase II development plans we previously announced and provide more details of those. So I'll also detail where each of RP2 and RP3 are going to be used for the program, including in the clinical trials where we're collaborating with Roche as we also announced earlier today. So as previously indicated, we're intending to start 3 Phase II clinical trials around mid next year with the first of these in head and neck cancer. We're here. We'll use RP3 and nivolumab under our collaboration with BMS. We do think that head and neck cancer may be a particularly low-hanging fruit indication for RP3, so that strays a little bit from our liver-centric approach, but we think there's a key opportunity in head and neck cancer for us beyond CSCC, which is often predominantly a head and neck disease as well. So this head and neck cancer trial has 2 cohorts. The first of these will be in patients with locally advanced disease where we'll combine RP3 with standard of care chemo and radiation and we'll then follow that with adjuvant nivolumab. This cohort is going to be randomized from the outset. So it's a relatively small -- large sorry, signal finding Phase II trial, which will recruit 100 patients randomized 1:1 to the standard of care compared to the standard care combined with RP3. We'll have an initial output of the metabolic response rate to the initial therapy by PET scan, and then comparative 1-year relapse-free survival as the key endpoint from that trial. So the intent of this trial is that it will evolve into a full registrational trial based on the initial data, which we hope to see. The second cohort in that clinical trial is going to be a 30-patient signal finding group of patients with recurrent or metastatic disease who have low PD-L1 levels. And it's those patients who do particularly poorly to standard of care chemotherapy or combination with anti-PD-1 therapy. So we'll focus on those low PD-L1 patients for signal-finding purposes. So here, we'll treat patients with nivolumab combined with chemotherapy and RP3, so essentially adding RP3 to the standard of care where we aim to see a good rate of response and duration of response. And then depending on that, we would intend to enter into registrational development. So the next 2 trials towards the bottom of the slide are under newly announced collaboration with Roche. So here, we will combine RP3 in both cases with atezolizumab and bevacizumab in both hepatocellular carcinoma and colorectal cancer. With one additional cohort in colorectal cancer using RP2 as compared to RP3. So each of these trials does have a faster market potential as there's no current standard of care, particularly in second-line hepatocellular carcinoma and third-line colorectal cancer. So as I said there, there really isn't a good standard of care. So one could imagine that single-arm registration development could be possible if we see a sufficient signal in the initial signal-finding cohorts. So these specific trials, as I said, will be in first-line and second-line hepatocellular carcinoma with RP3 combined with atezolizumab and bevacizumab in two 30-patient signal finding cohorts in third-line CRC, where 230 patient cohorts will also be enrolled. -- here, 1 with RP2 combined with the Atezo Bev and 1 with RP3 also combined with Atezo Bev. So for CRC, which is a particularly immunologically cold tumor type and particularly challenging to treat with immunotherapy, we do think this may be a particularly good setting to assess any differences between RP2 and RP3, although as you can see, most of the activity in this Phase II program will be conducted with RP3, which is our most advanced and intended and expected to be the most potent of our 3 RP viruses to date. So as you can see, these last 2 trials do fall right in line with our liver-centric strategy where most of the patients would be expected to have a liver-focused disease. So the next slide highlights both the challenges and the opportunity in patients with liver metastases. This includes that while the liver is the most common site of metastatic disease across really all tumor types, those patients also have a particularly poor prognosis. Liver metastases appear to specifically eliminate T cells from the circulation through the action of liver resident macrophages. This reduces the efficacy of anti-PD-1 based immunotherapy, which does rely on pre-existing T cells for its activity. So this is demonstrated on the right-hand side of the slide. where a bispecific anti-CTLA-4 and anti-PD-1 antibody demonstrated promising data with a 24% overall response rate in colorectal cancer. But notably, there were no responses at all in patients with liver metastases really highlighting the unmet need there. So RP viruses across our portfolio are designed to directly kill tumors, which remove the resident macrophage, [ saying, guys ], you kill the tumor, you also remove the macrophages and also to stimulate a systemic antitumor immune response, which should provide abundant new tumor-specific T cells to provide both systemic and durable effects. So we do think our new data with RP1 today in actual fact, further highlights that we -- the possibility of being able to do this today in PD-1 failed melanoma, including in the liver, which we, again -- does think provides us confidence that in these Phase II settings, we have a good chance of providing benefit. So in the next slide, we have -- this is a slide describing the experience we have so far in injecting patients in the liver, in particular, in relation to safety. So we have already, as you can see here, demonstrated that injections into the liver are both feasible and appear to be safe in patients with liver mets with the demonstration of clinical activity in a smallish number of patients so far in each of anti-PD-1 failed cutaneous melanoma, uveal melanoma, as Professor Harrington showed you earlier and also some patients with esophageal cancer and also MSI-high colorectal cancer. So you can see from the table that patients have received up to 8 doses, which is a full course of RP viruses injected into the liver with the median number of injections being either 5 or 6 depending on whether that was RP2 or RP3 in this data presentation, which was presented at a conference earlier in the year. So the side effect profile has been very similar to patients receiving injections at other sites, i.e., not in the liver, although there was a somewhat increased level of the same grade 1 and 2 side effects largely constitutional as Professor Middleton showed you earlier. So really the same 1 -- grade 1/2 profile, but possibly a little bit more frequent in patients following injections into the liver. So we think we've shown that injection of our viruses into liver is both feasible and appears to be safe and we've generated a strong initial demonstration of clinical activity in hard-to-treat tumor types, which we do think sets us up well for the Phase II program planned with particularly RP3 but to a lesser extent, RP2. So if we now move back to melanoma, in summary, we believe that RP1 is positioned from the day to day, combined with our prior data to really be a go-to treatment option for melanoma patients who progressed on or indeed after anti-PD-1 therapy with the data today confirming the signal we saw in our prior data. with, as you saw, a 36% overall response rate and a very impressive 20% complete response rate with, while the data is relatively early, good durability to date. Importantly, clinically meaningful activity has been seen across each of the settings in which patients need additional therapy following progression on anti-PD-1, both alone or indeed in combination with other agents, including anti-CTLA-4. And this includes in all stages of disease, including in patients with the most advanced stage 4 M1b and M1C disease, and includes a very striking 30% complete response rate and 50% overall response rate in the growing and rather challenging-to-treat segments of patients who progressed often rapidly while on adjuvant anti-PD-1 therapy. We also believe that bearing in mind the attractive safety profile we've seen so far and continue to show and the level of activity seen across a range of different anti-PD-1 failed melanoma settings, the RP1 may actually be optimally positioned directly after a patient's first exposure to anti-PD-1 including before they might receive, for example, an anti-CTLA-4 or other potentially toxic therapy where they haven't already had those as their first line therapy option. So with that, summary of -- further summary of the melanoma data, I'll hand back to Philip to wrap up.

Thanks, Rob. So before I wrap up, I actually want to speak to Rob's final remarks and remind everyone we've always said from the get-go, our modality and products can result in high rates of complete response. And we're seeing this play out in patients with end-stage disease in cutaneous squamous cell carcinoma and melanoma. These responses that transform lives give patients the potential for cure, and relieve them of disabilities and/or disfigurements caused by disease. We continue to believe that if our well-tolerated products are pushed earlier into disease courses, the rate of complete response can go higher still, and result in many patients never developing the type of end-stage disease we are currently treating. So on the alternate side, in summary, with RP1, we are on course to establish a major skin cancer franchise. In addition to today's melanoma update, we have completed accrual in our registration-directed randomized controlled CERPASS study in cutaneous squamous cell carcinoma and plan to release top line data in the first half of '23. Large scale manufacturing has been established and commercial planning for launch in the U.S. is advancing. With RP2/3, we have an exciting mid-stage programs pursuing GI cancers and head and neck cancer where expedited approvals in some settings could be feasible. We are pleased to announce a cost track collaboration in GI cancer system with Hoffmann-La Roche. Finally, we have a strong cash position to drive value through multiple meaningful data, near-term data catalysts. We will now turn over to Q&A. [Operator instructions] Thanks again for everyone's time this morning, and we now look forward to answering your questions.

Could please the management team and physicians turn their cameras on?

Thanks Philip, just to go to our first question. Were there any notable baseline differences between responders and nonresponders of RP1 plus OPDIVO, for example, timing of anti-PD-1 therapy or size of tumor lesions?

Take that in the first instance, Rob.

Sure. I don't think there's any notable differences. We can't really hone in on who might respond and who won't respond from the data. It really probably relates to the pace of progression. You can see that a significant proportion of patients really do progress rapidly, and as shown by the PFS curve. But if they get through that first period out to 2 or 3 months, then they have a good chance of ultimate benefit. So it probably relates to the pace of progression on the prior anti-PD-1 therapy before they come on to our trial. So more than that, I don't think we can really say.

Congratulations on the progress. Can you provide insight into what durability looks like in M1b and M1c patients?

So again, if you look at the swimmer's plot, that tells you the durability and we don't see any lesser durability in M1b or M1c patients than we see in the other stages of disease. As indicated, only 4 patients so far have progressed out of the 27 responding patients, and that is across the range of different -- in the overall population, not specifically M1b, c or any other stage.

Next question. Can you put the portion of adjuvant patient population enrolled in the study into the context of the real-world setting and put the response rate into context?

I think I might hand over to Dr. Mike Wong to address that, who has a large myeloma practice at MD Anderson.

Thank you, Dr. Coffin. I appreciate the opportunity to participate in this exciting data release. Patients who progress on adjuvant therapy in melanoma especially during the adjuvant phase are thought to be primarily refractory to immunotherapy. They have been traditionally a very difficult subset to treat. And oftentimes, we revert to combination checkpoint inhibition therapy recognizing this fact. I'm impressed by the fact that this particular population has a significant response rate. I believe it's 50% and a number of complete responders. To put things in perspective, this is a particular difficult patient population and to see a response in this setting is a signal that tells us that this is an active therapy.

So with regard to the proportion of patients in our group, do you see that as sort of representative of what you might see in your practice of patients needing for the therapy after anti-PD-1?

Yes, this is about how -- what we see in our patient population. Of course, it depends on whether you're a tertiary referral center or a primary center. But here at MD Anderson, because of our surgical practice, we see a number of these patients. And because of that, we have a large cohort of Stage 3 patients that are undergoing adjuvant therapy. And what we see here is representative of what we see in our patient population.

Thanks. Dr. Wong. The next question is, how will RP1 fit into the treatment landscaping, including TILs? And how will clinicians view the competing modalities?

Well, I think on that question, it would be good to hear from all of the physicians. So let me start with Dr. Wong and then go to Dr. Middleton and then Dr. Harrington.

I think that it's a true fact that the PD-1 refractory patients are difficult subset. There are -- and I see 2 things that are exciting to me. First, I talked about the primary refractory disease patients, which are those that failed adjuvant therapy. And also, I'm impressed by the fact that there are a significant proportion of patients on this cohort that we described, who have failed ipilitimab plus nivolumab. And these are patients who have received the top line most aggressive immunotherapy that we have currently that's FDA approved, and you're seeing significant activity in this subset. Again, it's -- it tells you there's a signal of activity that is significant here because that is a very difficult -- patients opted to treat because of this refractory nature of the checkpoints. There are no real good standard of care. So those are flags of activity for me. And when you see activity in this situation in the refractory situation, what that brings into the discussion is what role this particular strategy can play in a frontline setting. This has been mentioned before in this broadcast. And because of the safety profile, I can envision a situation where we can even think of neoadjuvant therapy, which is the hot area in skin cancers and melanoma presently. But I think this -- what I see here with data-wise really has legs. And I'll defer to my colleagues on the call, but I was formerly head of solid tumors and had a large sarcoma practice prior to coming MD Anderson, and I'm very impressed by the sarcoma data, which is traditionally known as a cold tumor. I'll stop here and let my colleagues weigh in.

So in the U.K., I think there's a significant concern about cellular therapies and their scalability. We operate in a more cost constrained model in the U.S. And when I look at what cell therapy is doing other tumor types, it's likely that PD-1 failed melanoma is going to get squeezed out, or will have severe limitations on what we can offer. And leaving aside the reimbursement issue, we also have to recognize the significant proportion of our patients who are relatively old, relatively comorbid and therefore, an option such as this with a similar response rate with a toxicity profile that analogous to having nivolumab or pembrolizumab as a single agent is going to be much more attractive. And just by way of example, the rate of primary single-agent PD-1 in the metastatic setting of melanoma was somewhat higher in Europe and in the U.K. than they are in the U.S., where that balance of toxicity and response perhaps puts more weight on the toxicity than is the case in the U.S.

Maybe I could just add a comment as well, Rob, on this. And I think the question I was asking really about I think attitudes maybe to sequencing treatments because, of course, for patients with this sort of problem, it's likely that they may need more than one therapeutic approach to this. So for me, one of the interesting things around intratumoral oncolytic immunotherapy is that it can potentially recondition that tumor microenvironment. And one could envisage a situation where, of course, in order to generate something like TIL, you have to resect the tumor, making it, therefore, unavailable for injection. One might envisage in the way that Dr. Professor Middleton and Professor Wong have said, you might actually see that injecting the tumor and giving rechallenge with anti-PD-1 therapy might be a very good first option. And if you then come to TIL therapy, you might actually be harvesting from a tumor deposit, where you may actually have reconditioned some of the micro environmental factors and who knows, it would be very interesting to see the quality of the TILs that come out of a lesion that has received oncolytic immunotherapy versus one that hasn't. So I can see lots of further opportunities there, which is, I guess, a little speculative. But I think I would see this as probably the viral therapy first and then the TIL therapy possibly is the backup, especially for those patients with more comorbid conditions.

Can we look back actually to Dr. Wong because it was sort of postulated that maybe in the U.S. we're prepared to see a slightly higher toxicity profile. So in trying to put that one back to you, Dr. Wong in terms of positioning our approach versus potential other approaches, including TILs as far as toxicity is concerned?

Sure. I mean, first of all, I -- this data speaks for itself, the paucity of grade 4 toxicities and an extremely low rate of grade 3 toxicities opens the door for a population that extends into geriatric range. That's important. Dr. Harrington actually has mentioned one of my favorite sort of postulates, which is if you have an educated conditioned immune system, that allows checkpoints to work better. And I remind folks that theoretically checkpoints, all we do is take the foot off the brake. And if we don't have a nascent antitumor immune response, then you don't get any response to whatever. And this is one of the mechanisms of why checkpoints do not work. And so I agree with Dr. Harrington that using oncolytics as a methodology to induce a primary novel immune response. It's a vaccination after all. And so everything that happens immunologically downstream of that is affected by it, including using combination checkpoints, including anti-LAG3, which are all checkpoints. So mechanistically, it is a completely different approach and the only one which allows us to generate a novel immune response. So I'm totally aligned with that strategy. In fact, this one of my teaching points when they give my immunotherapy lectures. Having said that, TIL is something we do at my institution. It's -- it's one of the leading institutions that does TIL, and we've been doing TIL for decades. I take no credit for it, many of my predecessors have really set the program on track, and we were also one of the clinical trial sites for some of the TIL products, which are in front of the FDA presently. It is an elaborate procedure. You have to have a proper patient that can have -- undergo surgery or resection of the tumor, then you have to have a lab component, which [ grows in mix legally ]. And in that interim time, the patient may or may not be on therapy. So there is the delay getting patients on. When you come into the hospital for TIL therapy, you have to undergo at least 5 days of top line chemotherapy to drive your immune system to 0, followed by infusion of these cells followed by [ hydros and leukin tube ]. Of course, there are different strategies going on, but that's sort of the backbone. And just by saying that, you can imagine that's not every patient, especially those who are in a geriatric range have any comorbidities. So despite our higher tolerance for I-O toxicity, we are highly selective about this resource intensive, both on the patient side and on the institution side therapy. So having something with a low morbidity can come in front of that makes a lot of sense, especially like to circle back to the statements made before, which especially if you can condition an environment which ensures even a higher rate of success.

Thank you, Dr. Wong. Our next question for the IGNYTE cohort, can you clarify timing expectations for enrollment completion? And how many months of follow-up you need for the full analysis?

This is a question for me yes. So the cohort is a 125-patient cohort in total. The data today was from the first 75 patients who have been followed for at least 6 months. So it was at least 6 months ago when we enrolled the 75th patient. We expect to complete the 125-patient enrollment around the end of the year -- very close to the end of the year. There will probably be a little bit of over-enrollment up to probably 130 something, which is similar to the CERPASS trial, which ultimately enrolled 211 patients, which is a little bit beyond our target enrollment. So in the first -- early part of next year, first couple of weeks to months, we should fully complete enrollment of this particular cohort. The primary analysis is triggered 1 year after the final patient has been enrolled. However, on the basis of this good data, we will be obviously discussing with the FDA next steps, including the design of a confirmatory study, which we will need at which point, we will also be sharing the data with the FDA to see exactly what the path forward with the agreement of the FDA is going to be.

Thanks, Rob. Next question, can you put into context the toxicity profile in light of the efficacy shown today?

I think the physicians have already obviously talked in relation to TILs, but maybe we could then also explore our toxicity profile, maybe any other options patients may have. I mean clearly, we have many patients have failed both CTLA-4 and anti-PD-1 and have very limited options, but maybe we could pose that question in terms of maybe patients who have only just had anti-PD-1 would then ipi/nivo be the treatment of choice in terms of potential safety efficacy promo as compared to RP1 on the basis of today's data, maybe the physicians could answer that question again with the same order, starting with Dr. Wong.

First of frontline ipilimumab plus nivolumab carries with it a significant response rate, but also a very significant toxicity rate. The original papers and what we've seen as well show that your high-grade toxicity rate grade 2, grade 4s are in the range of 50-plus up to 57% in trials. In original clinical trial, over half the patients could not complete all 4 cycles that were planned. The type of toxicities that we see can be significant including colitis, pneumonitis, myocarditis, and these are significant life-altering and oftentimes permanent issues. You see things like hypophysitis. When it occurred, I had to go and look it up at a book and turn to a medicine book. That's where you basically inflame your pituitary gland, you lose all your thyroid, adrenal adeno and sex hormone access. It's a significant issue. It impacts quality of life and also fertility. I'll stop here because it's just from these few seconds, you can understand, this is a very sort of powerful but particularly morbid situation. There is a foot dose, which a Phase II study have shown that you can actually use a different dose by altering the doses of ipilimumab and nivolumab still you may drop that high-grade toxicity rate into the 30% range, but the type of toxicities does not change. And then, of course, I already mentioned the -- what we have seen already on this presentation, the paucity of grade 3 toxicities and extremely low single-digit if at all, Grade 3 toxicities. I think the data speaks for itself.

Mark?

Yes. So rather than comment on standard care options, where I think we've got well understood inferior toxicity profiles for response rates that were certainly no better than our early data presented here today have shown, my practice is almost totally experimentally melanoma and almost exclusively PD-1 failed patients. So I have nothing on the books in the last few years or indeed presently, that has the same response rate and the same lack of toxicity. So if you're looking at the risk-benefit consideration, this is the -- I can't say market leader, it was just not on the market, but you know what I mean in the vernacular.

Any further comments, Kevin?

Yes. Maybe I could again just echo some of the comments of Mark Middleton. So our focus has been quite strongly on intratumoral therapies. And so I would look at the potential comparative toxicities with maybe other oncolytic agents that we've worked with and also the STING agonists because we have worked now with 3 STING agonist stages, all of which have thus far been delivered by intratumoral injections. And the toxicity profile of those agents is -- has been really quite impressive compared with very modest toxicity from RP1, modest, but possibly slightly more exaggerated toxicity with RP2. And I think it's probably fair to say, albeit with relatively limited data, RP3 with its greater cargo carriage is associated with potentially perhaps a greater cytokine release profile than maybe some of the other agents. But when we compare them and I compare them with my experience of using cyclic dinucleotide STING agonism, it is a milder toxicity burden. And most importantly, as the others have said, it's associated with responses which actually we haven't particularly been able to see in a strong way for the STING agonist. So I think you get the benefits of a modest toxicity profile for significant efficacy with this platform. And that for me is what makes it, again, avoiding Mark's term of market leader. But certainly, in terms of the profile of things we've been looking at, I haven't seen anything with this level of activity and tolerability in the last 3 to 5 years.

Thanks for our next question. Were there any patients that have received RP1 again at a later time point on disease recurrence where you saw a rescue of response?

Yes, we certainly saw that in our prior group of 16 patients, which we've reported on before, we've obviously got a longer follow-up than this group. So as a small minority did indeed have a reinitiation of RP1, both in the setting of patients who had extended stable disease but didn't quite get into response, patients who had extended stable disease and then began to progress and patients who had actual response and then new lesion appeared. And in each of those settings, we've seen reinitiation of RP1 to be valuable. The experience of reinitiation in this new group is less at the moment because we have less follow-up. But there are a number of patients well less than 10% who've had a reinitiated course of RP1 where there has been evidence of benefit, including allowing patients who haven't actually got into response to get to a response following a few extra doses of RP1, so we do think it's sort of a unique attribute of this type of therapy that you can reinitiate without expecting resistance to have kicked in. So you can imagine that actually patients could continue until futility kicked in to inject new lesions if they appear down the line, but that's very much in a minority of patients. Most patients who had a response, their responses tend to be durable. Mark and Kevin have both reinitiated patients with RP1 so may have further comments.

So one of the patients in the 16 the 75 cohort that I reinitiated led to a reduction essentially of a complete response after the reinitiation of RP1. She's now completed all treatment has just done the followup and remains disease-free. I think this falls into line with what we've learned around IO in general, which is that when we see lesion responding differently such as a new lesion responding on the background of overall success, then treating that aggressively has benefits. And whether that's vibration, excision, irradiation, or handily, the ability to reinitiate treatment with oncolytic virus, it has the potential for dividends. And clearly, in the background setting of overall success for the combination treatment, it's very attractive to be able to come again with what has worked before.

Our next question is how are you thinking about the commercialization of RP1 and the opportunity? And are there any worries resistant points from physicians, training, et cetera?

No, I think I sort of shared with the audience the fact that we're looking to launch broadly in the community, and we've been doing a lot of work around that, obviously learning some of the lessons from other products such as [ TLEC ]. One big thing is how can we just make this easy and routine for customers. And one aspect of that is how do we avoid some of the cold storage challenges as we think about communities, some of those practices don't have minus 70, 80 fridges. So if we can do extended storage at 2 to 8 degrees, which is regular refrigeration temperature that's one thing. We also have a very good safety profile. And in reality, oncolytic viruses are very safe to administer, but there can be institutional guidelines around certain things that add certain hurdles, and we're working actively to generate data working with institutional boards and other groups who can help us really think about what data do we need to generate to sort of overcome some of those hurdles. There's nothing there that we don't think we can overcome with data and education. And as we think about community practices, we really are feeling quite confident that this can be done. What we've heard drives adoption is having an option that addresses a high unmet need. And I think we've seen the data today. They clearly are having an unmet need for these patients often community practices will refer refractory patients for clinical trials or to academic centers because they don't want to deal with some of the toxicities, whether that's ipi/nivo and they certainly won't be able to handle TIL therapy. So we haven't heard anything. It's going to be a lot of education and whenever you introduce a new modality, that's part of the deal, but we've done a lot of work and we feel that it starts with strong data and working on some of those things that just make things more simple routine to incorporate into existing clinical guidelines and SOPs.

Thanks, Sush. The next question, were all of the PRs and CRs confirmed? So do you have any clue on how many intratumoral injections may be needed to drive a clinical outcome?

So yes, all the responses presented today are confirmed responses, and as you can see, nearly all of them all but 4 are durable to date, and with regard to a number of injections, the standard first treatment course with RP1 is up to 8 injections. And the median number of injections is either 7 or 8 in the patients who have responded. So most patients do get their full course of injections who are responding, patients actually across the whole group of patients, the median number was 7.

Next question is, how are you thinking about the potential to expand the development plan for RP1 in melanoma settings? You want to take that, Rob?

Yes. So we're in the relatively early stages of our thinking process there. And it does -- there is -- it is related to what we finally do as a confirmatory trial in melanoma. Obviously, we know that this is a single-arm data set and is expected and the FDA has informed us so the confirmatory trial would ultimately be required. However, optimally, we would conduct a confirmatory trial, which adds to the label, if we can, rather than just confirms the preexisting label. So exactly what we do in melanoma will be influenced by what we agree with the FDA to be a confirmatory trial, which we aim to discuss with the FDA in the early part of next year. But obviously, of particular relevance could be development in the neoadjuvant setting. I think our data very much speaks to development in the neoadjuvant setting. So while we don't have any firm plans currently to do so, it is certainly an area of great interest for us as it also is in cutaneous squamous cell carcinoma, where our thoughts are a little bit more advanced than in melanoma. So I think the conversations with the FDA in the first part of next year will really help us there. There are also other opportunities in melanoma as well, including things like mucosal melanoma, which are underserved currently and where there's nothing specifically approved. So again, while we don't have any formal plans at the moment, it's certainly also something on our radar for the future.

Thanks Robert. And this is a bit of a follow-up to that question. The data in uveal melanoma and sarcoma are also compelling, obviously, not the focus but assume the Phase I/II studies will remain ongoing and provide a basis for potential discussions with the FDA about the path forward.

So with regard to the uveal melanoma group with RP2, we have finished enrolling that. So we're not going to have more patients with uveal melanoma from that trial, but we do think already that, that data is very strong. And as I said, well, again, we don't have any formal plans at the moment to proceed with further development in uveal melanoma, it's certainly something very solidly on our radar and which we're thinking about. There will be some patients with RP3 also treated with uveal melanoma, so we'll get a feel for RP3 in uveal melanoma too before formally deciding what to do. In sarcoma, really, the situation is very similar. We think we have a good strong initial signal worthy of further development. And as a company, we're sort of huddling together to think about exactly how we should take that forward. in the context of everything else we're already doing, including preparing for the Phase II trials in head and neck, HCC and CRC. So I would say watch this space would be the real answer to that.

The next question, did you observe differences in clinical data or biomarkers that led you to choose RP3 over RP2 or just on the expectations that RP3 should be more potent and RP3 safety was comparable to RP2?

To some extent, the latter in reality. Obviously, we've only treated a relatively small number of patients with both RP2 and RP3 and trying to definitively determine which "may be better" is challenging where both of the products are clearly clinically active. So the reason for proceeding with RP3 in head and neck and mainly in HCC and CRC included that we have seen good safety, although a little bit elevated side effects, which are on-target. So we think that's good, but no less tolerable combined with we have seen clear evidence that RP3 is clinically active and theoretically, and preclinically, should be and is more active. And for that reason, we have with our partners determined that we'll largely proceed with RP3 while still conducting an experiment comparing RP2 and RP3 in the particularly challenging situation of colorectal cancer, but we expect that RP3 will pan out to be at least as active and potentially more active.

And can you maybe a bit of an add-on to that, Rob, but can you provide any additional color on your decision to move forward with RP3 and head and neck and HCC and to access RP2 for CRC?

Well, I think I'll refer you to my prior comment. Professor Harrington will be or one of the co-PIs on the head and neck cancer study, and maybe Kevin may want to comment on the head and neck study in RP3, using RP3 for it?

Yes. Thanks, Rob. So I think you could probably -- we could probably make a good case. And I apologize for a bit of background noise. I'm just on the periphery of the poster session here that's just starting. But you could make a case perhaps to the RP2 agent, but for me, I think the RP3 agent in head and neck cancer brings a great deal to the party. So I think from my perspective, the role of the CD40 ligand and the 4-1BB ligand agonism, both of which, as agents have shown some anecdotal evidence of impressive responses in cohorts of patients with head and neck cancer without necessarily themselves justifying registrational path type studies, the expression from those agents from the virus is particularly attractive. The potential role also of CTLA-4 blockade in head and neck cancer, albeit with the negative Phase III studies, CheckMate-651 that we were involved in, 714 also and also the study with KESTREL that's about to be presented although those were headline negative for CTLA-4, there are potentially evidence -- there is potentially evidence of really impressive responses for some patients. And I just draw your attention, for instance, to the CheckMate-651 showing that the median overall survival in the first-line setting with combination of ipilimumab and nivolumab was 18 months, which is unprecedented in patients with relapsed head and neck cancer in the first-line setting. So for me, the package that's in RP3 looks particularly attractive for patients with head and neck cancer. But then again, I think you could probably make a case for RP2, but you can only do so many studies. And for me, RP3 is more attractive. And again, I apologize for background noise.

As a reminder, we'll be in both RP2 and our P3 in colorectal cancer.

And we have time for one last question. So perhaps we'll turn it back to the IGNYTE data. And Philip, do you expect the final 125-patient data to be similar to the first -- the data you've seen in the first 75?

We can't say for sure, exactly. But what I can say, if you look at the overall response rate table that we see very strong rates of activity in all subtypes, including late stage and/or M1b disease with a 28% response rate, including a non-adjuvant with a 30% response rate. So it's hard to see whatever the final mix, the overall picture is going to change in terms of the fact that we -- with the side effect profile, efficacy profile that we presented to date, we will not have potentially here a product for all melanoma presentations.

And that concludes question and answer. If we did not get to your question, we will do our best to get back to you. We did have a lot, so I apologize for that, but I'll turn it back to Philip for closing remarks.

Yes. No, I'd just like to thank those for listening in. I'd particularly like to thank the physicians for joining us on our panel, Drs. Wong, Middleton and Harrington. And I'd also actually like to thank our staff, both here in Woburn, [ framing ] and the U.K. for their dedication to the course. Thank you very much.