Welcome to the 41st Annual JP Morgan Healthcare Conference. My name's Anupam Rama. I'm one of the Senior Biotech Analysts here at JP Morgan. I am joined by my team, Priyanka Grover and Malcolm Kuno. Our next presenting company is Replimune, and presenting on behalf of the company, we have the CEO, Philip. Thanks, Anupam. Replimune. Replimune is the industry leader in tumor-directed oncolytic immunotherapy. We have three wholly owned programs, RP1, 2, and 3. RP1, day-to-date, supports the emergence of a major skin cancer franchise. We have two registrational studies ongoing. The first is in front line cutaneous squamous cell carcinoma, where in a prior single-arm study, we showed a very impressive almost one in two complete response rate. We have subsequently fully enrolled a 211 patient randomized controlled study of RP1 plus cemiplimab standard of care against cemiplimab alone. We look forward to disclosing top-line data from this pivotal study later this half. The second study we're running with registration intent is in the anti-PD-1 failed melanoma setting. Here, there is no standard of care. We're running a single-arm study. We disclosed data from the first 75 patients late last year with a very impressive 20% complete response rate and a 36% overall response rate. We have a broad mid-stage development plan with RP2/3. Our pipeline RP2/3 targets less immune sensitive tumor types, where we've seen robust activity in phase I, and we plan to start the phase II program midyear in head and neck cancer, HCC, and CRC. In third line CRC and second line HCC, we also recently announced a cost-sharing collaboration with Roche. As such, across our portfolio, we have the potential to deliver substantial revenues beginning in 2025. We're capitalized to build a fully integrated biotech company with over $600 million on the balance sheet. We've bought manufacturing in-house and are completing our process validation runs for our BLA. We continue to strengthen our commercial infrastructure with the recent appointment of Chris Sarchi as Chief Commercial Officer, who, while running the commercial oncology franchise at Sanofi, launched cemiplimab as the current standard of care in our lead indication of cutaneous squamous cell carcinoma. Just a reminder what oncolytic immunotherapy is, the use of viruses that selectively replicate in and kill tumor cells, but do not replicate in healthy tissue and leave them unharmed. On injection into a tumor deposit, that tumor deposit is invariably either completely or partially destroyed. The tumor cells are lysed open by the virus, and the cancer antigens within are exposed to the immune system in an environment of necrotic cell death. This attracts antigen presenting cells to the injection site. They internalize those escaped cancer antigens, drain to the lymph nodes, and prime T-cells to destroy uninjected deposits throughout the body. Dual mechanism of action, direct tumor-mediated viral cell lysis from the engendering of a full systemic immune response. Our platform is based on the herpes virus, which we think is an ideal viral species for oncolytic use, it being naturally highly lytic and having high carrying capacity to carry in multiple other immune-stimulating proteins into the tumor microenvironment to further amplify the immune response. Our herpes strain is based on a new clinical strain, specifically selected for its lytic properties. From all of our products and part of our platform, we express a glycoprotein, which further increases the lytic potential and immunogenic potential of construct 10-100 fold. This is essentially RP1, where we plan to establish a major skin cancer franchise. RP2 additionally expresses an anti-CTLA-4 antibody, RP3 additionally expresses ligands targeting CD40 and 4-1BB. As I said before, here we're targeting less immune sensitive tumor types with a phase II program to commence midyear in head and neck cancer, CRC, and HCC. Focusing on RP1 and the establishment of a broad skin cancer franchise, the two main planks of which are these two registrational studies in PD-1 failed melanoma and first-line cutaneous squamous cell carcinoma. We've also seen activity in other tumor types, including Merkel, basal, angiosarcoma, both in the first-line setting and in the PD-1 refractory setting, which we could potentially add as compendia listing. We have our ARTACUS study in patients who have developed skin cancers after being immunosuppressed for organ graft transplant. Here, PD-1 is not typically used due to risk of loss of graft, and we're using RP1 as monotherapy. We've seen activity. We look forward to further updating the street on this study throughout the course of this year. The setting where we generated the most data to date that we have disclosed is in the PD-1 failed melanoma setting, 91 patients in total, 16 patients from a phase I/II study, and then the first 75 patients from the 125 patient study we're running with registrational intent, with an impressive 36% response rate to date. Of note, the majority is responding patients with primary refractory anti-PD-1. They did not respond to a first line of anti-PD-1 therapy. Further anti-PD-1 use is deemed futile using the strict SITC criteria. What is also pleasing is that we have seen robust activity in all presentations of the disease in the refractory setting, including late stage IV M1B, M1C visceral patients with a 28% response rate. In patients who have not just failed PD-1, but failed CTLA-4 with an almost 30% response rate. Then in adjuvant progressors, these are patients that have had surgery and been rendered no avoidable disease and given PD-1 to stop the disease coming back. They have progressed, and in some case, with substantial disease burden, we've seen a very impressive 50% response rate and a 30% complete response rate. Pictures tell a thousand words. This is the waterfall plot here showing that in fact, over 50% of our study population has benefited. It shows the depth of response. Over half of our responses are, in fact, complete ones, giving patients a chance for cure. This includes patients with end stage four, M1B, M1C disease, where we've also generated complete responses. The swimmers plot here shows durability. The blue circle signifies PD. We only have four of those. 85% of our responses are still ongoing, up to the longest time point of over 1,000 days. 60% almost of our responders are already out to over one year. Even though this is an immature data set, this number will go higher still. Let's see some patient examples here. This is actually rather atypical. This patient did respond to a first line of OPDIVO, then progressed, and was then actually given KEYTRUDA, which did not respond to at all, while progressing extensively in the viscera when coming onto study. Injection into this forehead mass, which resolved in the red circle, and then we show in yellow circles a resolution of uninjected disease after the T-cells have been primed to go off and destroy uninjected deposits. As you look from top to bottom, you can see resolution of disease, extensive disease in the liver, and in that right-hand panel, as you look from top to bottom, you see resolution disease in the hilum area of the lung. This was a patient who failed single agent KEYTRUDA and then failed combined anti-PD-1 CTLA-4 therapy before coming onto study, with extensive disease in the lung. Here we injected disease in the chest wall and the back signified in the red circles. Again, we've highlighted uninjected resolution of disease of the T-cell priming with the yellow circles. You can see resolution of disease there in the chest wall in yellow, and then this bulky deposit in the lung as you look from left to right, you can see resolution of that disease, as well. This study, we reached a target accrual of 125 patients to date, and we will give the update from the full study readout towards the end of the year. Our lead indication is in fact cutaneous squamous cell carcinoma, not as well known as melanoma, but is the second most common skin cancer and actually has as many deaths, as melanoma annually in the United States. It's a little bit atypical for cancers in that mortality is typically driven by locoregional progression rather than metastatic disease. It often being a disease of the head, neck, scalp region and invades vital structures. Up until 2018, there was really no effective therapy for this disease. A PD-1 in the form of cemiplimab, Regeneron's anti-PD-1, was approved in 2018 and is an effective drug with a 35%-45% response rate, but only a complete response rate in the 5%-15% range. We showed in a prior single-arm study with Nivo a very impressive complete response rate of almost 1 in 2 patients of around about 50%. Again, pictures tell a thousand words, best exemplified by this spider plot, where you can see the depth of response, patients going to complete response relatively rapidly and remaining in complete response for out to 800 days. We only have one progressor here that has gone to complete response, where we've actually reinitiated treatment and put the patient back into response. A couple of examples. Here's complete resolution of aggressive locoregional disease, an ulcerated foot mass, injections into inguinal nodes. You can see resolution of disease both in the foot and in the groin area. Here's an example of complete resolution of disease in a metastatic patient with multiple bulky neck deposits, a retroperitoneal node, and multiple deposits in the spine. This patient was declared a complete response at one year and remains in complete response three years later. We do believe that it is complete response that have the opportunity to transform the CSCC market if we can double or do better than that complete response over checkpoint blockade alone. As this KOL quote says, CRs are very important in this setting as they usually lead to long-term survival. In cancers like melanoma, it's been proven that CRs correlate very closely with survival, not so much Partial Response. Obviously, if you're a patient, you're gonna much prefer that all disease has resolved rather than you having residual disease burden. Our registrational randomized controlled studies, as I said at the beginning, is fully enrolled, 211 patients comparing RP1 plus the standard care cemiplimab against cemiplimab alone with dual independent primary endpoints of complete response and overall response. It's been powered to show a 15% absolute difference over the control arm for success. If you take the complete response rate historically shown with cemiplimab, the higher end of the 5%-15% range, 15%, and add a 15% delta, we would win if we get a 30% complete response rate in this study. As we've seen in the CERPASS study, we in fact have nearer a 50% complete response rate, so quite some margin for error. In terms of overall response, the historical range, 35%-45% with cemiplimab alone, adding 15% to the higher end of that range means that we'd win at around 60%, and we saw a 65% response rate in the prior single arm study. Not so much margin for error, but I would remind you that these are independent primary endpoints. We split the alpha and we only have to win on one. In terms of commercial opportunity, there are about 11,000 first-line cutaneous squamous cell carcinoma patients treated in the U.S. every year. We do believe if we can double or do better the complete response rate, that we should be able to actually grow the market and there's a potential for another 10,000 patients to be treated. In anti-PD-1 failed melanoma, there are about 13,000 patients that fail PD-1 every year. I think if we go earlier into the disease course, particularly into the neoadjuvant setting, there are up to about 50,000 patients to potentially address. Our modality is very well-tolerated, only grade one, two constitutional symptoms, gives a high rate of complete response, and therefore, early in disease course makes a lot of sense for us in terms of future development plans. In terms of the number of patients that we can address with an intratumoral injection technique, cutaneous squamous cell carcinoma is effectively nearly all. Around 80%-90% can be addressed with superficial injections, more than that using ultrasound techniques. Using superficial injections and relatively straightforward ultrasound techniques, we can also address 60%-70% of the melanoma market, go higher still if we use other deep injection techniques. We do believe that it's feasible for this to be routinely incorporated across the majority of practice settings, including in the community. We do plan to launch in tandem in the community with academic and multidisciplinary centers. To do so, we're also making sure that the logistics is straightforward. We will have a fridge. Temperature presentation of our products will not rely on -80s, we're also going to address any misperceptions as they relate to biosafety. We're rolling out a robust publication plan. In terms of manufacturing, we've brought manufacturing in-house, our own facility in Framingham, both RP1, 2, and 3. RP1 has been released to clinic after we've shown comparability to contract materials used in our studies to date, and our BLA process validation consistency lot campaign is concluding. That's RP1. I'm moving on to our pipeline of RP2/3. RP2 is RP1 that additionally expresses an anti-CTLA-4 antibody, and RP3 additionally expresses ligands targeting CD40 and 4-1BB. We're trying to achieve here is maximum T-cell priming at the APC T-cell interface to basically overcome resistance in cold tumors and engender the most powerful immune response possible. We've shown robust single-agent activity with RP2 in what are typically very difficult tumors to treat in the salvage therapy setting, including mucoepidermoid carcinoma, which is entirely resistant to checkpoint approaches. Uveal melanoma, which very poorly responds to checkpoint approaches, and in any case, this patient had failed ipi/nivo. Also esophageal cancer in a patient that had failed anti-PD-L1, which is otherwise moderately receptive to checkpoint approaches. This is a CR in the salivary gland cancer patient, which remains ongoing to date out to over two years. This is the uveal melanoma example. These patients have eye cancer. The eye is typically removed, and 9% of the patients metastasize to the liver, and when they do, survival at one year is only around 10% or 15%. A very difficult disease to treat. This patient was injected in the large red mass, which you can see resolving away. You can see the priming of the uninjected deposits destroy the uninjected masses in the yellow circles. This is the esophageal cancer patient I mentioned, again, with injections into multiple tumors in the liver, which resolved away. You can see this yellow circle abdominal mass is no longer metabolically active. This patient is a complete metabolic response over 2 years from coming onto therapy, having failed 6 lines of prior therapy. Subsequent to phase I, we've developed a phase II program in head and neck cancer, primary liver cancer, HCC, and colorectal cancer. RP3 is designed to be the most potent of our products. We'll be trialing RP3 in all of these indications. We will be comparing RP2 to RP3 in CRC. As I mentioned at the beginning, in second line HCC and third line CRC, which are potentially faster market opportunities, we're doing a cost-sharing collaboration with Roche, which we announced in December. Why did we choose these indications? It's a combination of commercial potential, ease of administration, and scientific and clinical rationale. In terms of ease of administration, whether it's RP1, 2, or 3, all the tumor types we've gone after are either predominantly superficial in nature or are liver-centric, and it's fairly straightforward to inject the liver under ultrasound guidance. In terms of clinical and scientific rationale, squamous cell carcinoma of the head and the neck is a corollary to cutaneous squamous cell carcinoma, where we've seen very high complete response rates. HCC and CRC are liver-centric diseases, whereby we have chosen to go after this unmet need because of the anecdotes that I've shown you today, we have shown in multiple tumor types, we can inject liver deposits and see the injected tumor resolve, see uninjected tumors in the liver resolve, and see the scope of effects beyond the liver. It really is a really major unmet need, white space in oncology. Once patients metastasize to liver, their prognosis plummets, their survival odds plummet, and their response rates to checkpoint blockade drugs plummet as well. It doesn't matter whether it's melanoma, lung, renal, colon. Here's an example from Genentech's data in CRC. Very impressive 24% response rate in second-line CRC. Odd note, the patients that actually had liver involvement, liver metastases, the response rate was zero. A real unmet need. The hypothesis here is once a patient has liver mets, the macrophages within take out anticancer circulating T-cells from the bloodstream and switch to an immunosuppressive environment. With a powerful oncolytic, we can remove that sink, remove that tumor sink, we can splice open the tumor cells and release the cancer antigens to generate a systemic immune response to destroy uninjected deposits, at the same time, vaccinate against future relapse. We look forward to further trialing these studies and this hypothesis that we can have benefit here in both phase I and phase II studies during the course of this year. In summary, with RP1, a major skin cancer franchise is planned. We've generated strong data to date in multiple skin cancers and in the anti-PD-1 failed setting, we think the data is potentially transformative. We look forward to disclosing data from our cutaneous squamous cell carcinoma study, the CERPASS study. It's a registrational study later this half. Scale manufacturing in place and commercial plans are ramping up. With RP2/3, we will be commencing midyear our phase II program in head and neck cancer, CRC and HCC. We'll also have phase I updates, also during the course of the year. We have a very strong cash position which execute on our vision with $660 million as of the 31st of December, we have a catalyst-rich year ahead of us, including, as I aforementioned, the CERPASS primary readout, the readout from our transplant study, the ARTACUS study, whereby we are using RP1 as monotherapy. We'll have a 30-patient non-melanoma skin cancer study, which we fully enrolled leading out this year in the PD-1 failed setting. As I just mentioned, we will have further phase I data with RP23 and plan to get our RP23 phase II program underway midyear, such that we should have data within 2024, for which we have a substantial cash cushion beyond as well in terms of our cash outdate. Thank you. Thank you. Thank you, Philip. Just as a reminder, if there are microphones in the room, should you like to ask a question, just raise your hand and we'll make sure that happens. I don't think the ask-the-question feature works here, due to internet connectivity issues. Philip, thank you so much for the presentation. On the regulatory side for RP1 and PD-1 failed melanoma, what gives you confidence that a single-arm approach in IGNYTE is going to be accepted by regulators? How much duration of follow-up do you think you will be need there in a single-arm fashion? Want to answer that, Rob? Yeah, I'm Robert Coffin, and I'm in charge of the R&D at Replimune. We have had discussions with the FDA, particularly a Type B meeting with the FDA, nearly a couple of years ago now, where we discussed whether a single-arm study would be acceptable for accelerated approval. The feedback from the FDA is, was that as long as the data in its totality was sufficiently compelling, that it's certainly a setting where an accelerated approval filing could be accepted by the FDA. We have had no indication that anything has changed since, and have had to add more recent communications with the FDA, where indeed we haven't heard anything different. With regard to durability, various discussions with the FDA indicate that responses which last at least 6 months are deemed to be clinically meaningful as far as the FDA is concerned. A key secondary endpoint of the study is therefore duration of response. We do think from the data we've already shown, both in this most recent data cut and before, that our responses tend to be very durable, and therefore we do expect the vast majority of our responses to last for at least 6 months, and therefore easily meet that hurdle. I'd also think that the data we presented in December really does show a level of compellingness which meets the criteria for overall compellingness, which the FDA discussed with us at the Type B meeting. Questions from the audience? What are the plans in the EU on the regulatory side in anti-PD-1 failed melanoma? I don't think we're ready to actually say anything about that publicly. There are analysis ongoing, but obviously it's not just about getting approval in Europe, which is one hurdle. The real hurdle is getting reimbursement and making sure that your package supports that in the major European territories. We're not prepared to actually say anything more than that at this point while our analyses are ongoing. One of the interesting things is on the commercial side. How do we think about the peak potential in anti-PD-1 failed melanoma? I think for some of us, some of the KOL feedback has been that, well, not all of the lesions are injectable, right? That could limit the size of the market. How would you respond to that pushback? What are you doing to kind of educate physicians to sort of maybe maximize the potential of the indication? Sushil? I'll take the second part of your question first in terms of the superficial injections. In checkpoint failed melanoma, about 30%-40% of lesions are superficial and can be easily injected. If we wanna get to a broader population, that is gonna require some sort of ultrasound, but that's readily available and something that IRs do all the time. Community oncologists and others commonly work with the IR community to do things like biopsies, so we really don't think that's gonna be a major barrier. Certainly, we've also seen sort of impressive data in the stage four M1B and M1C patients, which is also gonna drive the adoption and sort of willingness to do those liver injections, sort of those deeper injections. On the other side, if you think about the peak potential, as Philip said, one of the things we're excited about is you think about uptake. There's two really areas is, you know, what's the patient opportunity? Based on the IGNYTE 75 patient data analysis, we really see that we're a great option for the range of anti-PD-1 failed presentations. The other thing that's really important in this setting, especially given if the other treatment options, is the tolerability profile, which I think in combination with Nivolumab is extremely tolerable. If you look at some of the other options in this space, the toxicity can be quite significant for patients. This is another opportunity to grow the market because some patients forego treatment given toxicity of existing treatments. I'll sort of add as well. I think it's very important to remember that we don't need to inject anything like all the lesions. We only really need to have at least one, you can inject with relative ease. In the data we presented in December, we really did show extremely compelling systemic overall responses in both injected and uninjected lesions, with the majority, the vast majority of responding patients having both injected and uninjected lesions. Only need to be able to get a needle really to one lesion, to be eligible for treatment. Yeah. You talked about what else we're gonna need to do around education. Just one thing to add to some of the KOL feedback we've had is right now, KOLs are not looking for lesions to inject, but obviously with this data, they said they'd actually are likely to find a lesion if to inject, given the sort of transformative data we've seen. In terms of other things we need to consider around education, there's really key, two or three key things, really. Identifying injector champions at each of the sites, We've done a lot of research around that. This will be advanced practice physicians, you know, APs, PAs are able to do this with a little bit more training. Certainly, experienced nurses can do this. It's not particularly difficult technically. If you've got people who have done extensive T-VEC experience, this is something that can be done in 10 or 15 minutes. As Philip said, we're gonna have to train a little bit on some of the biosafety issues. Again, this is more like a vaccine, and really we haven't seen any major safety concerns. I think this is more around education. Question from the audience. Terry. What do you think of the Moderna data? I don't think, and correct me if I'm wrong, I don't think you guys could do adjuvant, thoughts on going to perhaps neoadjuvant or something earlier? Yeah, we can't do adjuvant as we have to have something to inject, but obviously we can do neoadjuvant. Do you wanna take that, Will? I think the Moderna data is certainly very interesting. If there is one place where an RNA vaccine may really provide benefit in cancer, I think it's always gonna be in the adjuvant space, i.e., preventing relapse rather than actually getting rid of pre-existing disease. It's obviously not a direct competition for us 'cause as discussed, we would be interested in playing in the neoadjuvant space, but aren't able to play in the adjuvant space. I think the data indicates a compelling clinical activity. It's obviously early and small, but certainly worthwhile to build upon. I would add that there's no real evidence that it's actually better than giving neoadjuvant PD-1 followed by adjuvant PD-1. Thank you. What are the potential read-throughs from IGNYTE to the upcoming CERPASS data, for RP1 plus cemiplimab-CSCC? Wanna take that first? Sure. I think one of the very pleasing things about the data which we presented in December, was that it was a follow-on from prior data in only 16 patients, where we showed an essentially identical response rate. We had a 36% response rate in the, in the prior 16 patients, which led us to do the pivotal cohort in 125 patients. It's very pleasing that in the first 75 patients from that pivotal cohort, the response rate has been maintained at coincidentally exactly the same level. The CERPASS trial, which is a randomized controlled phase II trial, with registrational intent in CSCC, was likewise based on a much smaller amount of data in CSCC, again, just 17 patients, where we're obviously hoping to see similar results in 211 patients. I think the fact that the melanoma data was maintained between a small and a larger group, gives us stronger confidence that the data from that small group was in inverted commas, real, and not just anecdotes which wouldn't be repeated in a larger patient's number. Certainly adds to our internal confidence. I'd also add... What about the CR rate? Yeah. Yes. I would add that the fact that we've got, already got a 20% CR rate in unmet need population gives us even further confidence in terms of our ability for this modality to affect complete response. Moreover, the 36% response rate, essentially to show the response rate differential in the cemiplimab study, you're going to have to benefit patients that would otherwise be a primary factory to checkpoint blockade disease, and that's exactly what we've shown in the melanoma. That would give us additional confidence too. Questions from the audience. Do you see survival benefits? Obviously to date, our data has been single arm, and therefore we can only sort of wave our arms around in reality about survival at this point. We do think our survival curves are impressive and very impressive. However, they are single arm and therefore need to be taken with a pinch of salt. That's all I can really say. When you say you have an exploratory regulatory pathway, do you have to do confirmatory studies for survival? Like IMLYGIC clearly didn't show survival, right? I mean, it showed partial response. IMLYGIC showed a very strong survival benefit in half of the population, which was the population who didn't have visceral metastases. Actually, hazard ratio of 0.5 for survival in that group. Unfortunately for M1b and M1c disease, showed a hazard ratio of 1 for survival. I didn't work for those. Overall missed survival by one patient. As a result, we do expect to have to do a confirmatory trial in melanoma. It would be expected to be an accelerated approval, by definition therefore, we would need to do a confirmatory trial. One of the two regulatory interactions intended for the first six months of this year is to really agree with the FDA what the confirmatory trial may be, which again, by definition would need to be in a slightly different population to which we sought the accelerated approval, because if there were an appropriate control arm one could use in that group, we would've had one already. The only reason it's a single arm is there isn't an appropriate control arm to use. it would need to be in a situation where you can have a control arm, which as I said, we're intending to discuss with the FDA. Do you expect your confirmatory trials to be completely enrolled? I know regulations are changing with the... From a regulatory perspective? All of our feedback from the FDA is that as long as we have the confirmatory trial started when we file for accelerated approval in that indication, then that ticks the box as far as the FDA is concerned. We're dealing with CBER, not CDER. We've got no suggestion from our interactions that their view is different to that. You know, it's also a fact that Iovance's BLA has just been accepted for filing, and I don't think they started their confirmatory study, and that's a very similar setting. Going back to your previous question on survival, I would say that I think in a single-arm study, the best proxy you can possibly have for survival is durable, complete response. Our durable complete response rate in all of the settings that we disclosed is very impressive. Terry? Your first filing will be in CSCC, right? Which is for full approval, correct? Correct. Okay. Do we really need to worry about is the confirmatory study for the second indication underway if you're already approved, fully approved for something else? The confirmatory study would be to maintain the label for PD-1 failed melanoma, but would not be in any way relevant to CS, first-line CSCC. Got it. One last one on CERPASS. Will we perhaps get survival out of this? Is there a crossover at a certain point, or how long might this take in CSCC? There is no crossover in the design to enable us to collect survival data, although the study wasn't designed with survival in mind. We do expect over time, the survival to mature to show a survival benefit if we're having a true treatment benefit, particularly as we expect it to drive a much higher proportion of complete responses which are durable in the combination arm than the control arm. On CERPASS, how would you characterize the patient population being enrolled, relative to the cemiplimab pivotal study alone? How would you expect the monotherapy arm to perform? We don't have any particular reason to expect our population to be any different than prior populations in CSCC. Our inclusion criteria are the same or essentially the same. However, different studies with anti-PD-1s have given different results with Nivolumab, pembrolizumab, or cemiplimab of anywhere between around 35% and around 55%. While the inclusion criteria may be relatively similar, probably the exact breakdown of patients who happen to be enrolled with regard to severity of disease, probably has influenced that range of somewhere between 35% and 55% in those different, all of which are single-arm studies. We expect the control arm to come within somewhere in that range. As Philip indicated, to win, in inverted commas, on ORR, we need to see a 15% absolute difference between the arms, whatever is seen, which is the case for both ORR and CR. For CR, the historical data with different anti-PD-1s actually ranges from 0% to around 15%, depending on the PD-1. Likewise, we need to see an absolute 15% increase to have success on that other independent primary endpoint. To stress, we only need to win on one to have a positive trial. Our view is that CR is the most important because that's really what benefits patients more than PRs. Any final questions? Okay. Thank you, Philip and team.