Evaxion A/S (EVAX) Earnings Call Transcript & Summary

Good afternoon, and welcome to the Evaxion Biotech KOL Webinar on Evaxion's oncology vaccine. [Operator Instructions] As a reminder, this call is being recorded, and a replay will be made available on the Evaxion website following the conclusion of the event. I'd now like to turn the call over to your host, Dr. Erik Heegaard, Chief Medical Officer of Evaxion. Please go ahead, Erik.

Thanks, Tara. Hello, and welcome, everyone. My name is Erik Heegaard, and I serve as the Chief Medical Officer at Evaxion. I'm joined today by Niels Moeller, who is the Co-Founder and Board member. Evaxion is a clinical stage biotechnology company, specializing in the development of AI-based immunotherapies, and I'm delighted to welcome you all to this key opinion leader webinar on metastatic melanoma and personalized cancer immunotherapies. Next slide, please. So this presentation will contain forward-looking statements, as outlined in this slide. And next slide, please. And this is just the agenda at a glance. The webinar will feature a presentation from our key opinion leaders, Dr. Jeff Weber from NYU School of Medicine; and Dr. Patrick Ott from Dana Farber Cancer Institute of Boston. And specifically, Dr. Weber will discuss the current treatment landscape and unmet medical needs in patients with metastatic melanoma. Dr. Ott will then discuss the role and potential of personalized neoepitope-based cancer immunotherapies. And this -- following these presentations by Dr. Weber and Dr. Ott, Evaxion's management team, that will be Niels and I, will provide a company update and a live Q&A session will follow. Next slide, please. So based on our distinguish presenters, they have -- both our distinguish presenters have a very strong professional track record and are spearheading the continued development in the field of melanoma and immunotherapies. And I will let you dive into their many accomplishments by reading the attached bios at a later stage. So next slide is actually me handing it over to Jeff Weber. Jeff, please?

Thanks, and thanks for the invitation to make the presentation. As you indicated, I'll be giving you a quick overview of sort of the current state of the art of melanoma therapy, both metastatic and adjuvant therapy. And the current unmet need will be very clear because if one looks at the survival curves that I will show you, it's very clear that at the end of the day, half of all patients with metastatic melanoma and probably half of all patients with resected melanoma are still in need of new therapy. So go to the next slide. It's my disclosures that the lawyers require me to show this every time I speak anywhere, anytime, internal or external. Go to the next slide, please. Again, the study that really set the tone for what is generally the standard of care for most patients in the community who have metastatic melanoma, who is the so-called KEYNOTE-006 study. I'm not going to go all the way back to tell you about the very first Phase I trials. I'm just going to tell you about the trend setting and the registrational trials that have really set the standard of care. And KEYNOTE-006 was a large randomized study that for the first time showed that a PD-1 antibody, namely pembrolizumab, was clearly superior in terms of its overall survival to what was then the standard of care in the U.S. and the EU of ipilimumab, which was the first checkpoint inhibitor to be approved back in 2012. This was a rather large study over 800 patients, and patients got 1 of 2 schedules of pembrolizumab at 1 dose versus, again, the then standard of care for frontline melanoma therapy of ipilimumab, which was only given for 4 doses and then you were done. And it's an interesting study because if you responded, you were stable, partial complete responder. Of course, you were followed. But if you progressed, you were guaranteed getting retreated with pembrolizumab. So not only did you look at the frontline impact of pembro, but you could also look at the second line impact of repeating pembro. So if you go to the next slide, it shows the updated data that we just presented about 4 or 5 months ago by [ Caroline Robert ], who is the PI of the trial. And it shows that when you combine the 2 schedules of pembrolizumab together in the green line, it's pretty obvious that you have a superior survival to the standard ipilimumab in sort of the crimson line. But not only that, you have a plateau that goes on from 5 years to 6 years to 7 years, and this is a 7-year follow-up, a phenomenal number. And it looks like almost 40% of the patients are still alive and frankly, doing well. So you can -- with a single drug, pembrolizumab, you can probably cure at least 1/3 of the patients, maybe as many as 38%, 39% of the patients. The median survivals, of course, are very different. It's more than doubled with pembrolizumab. The hazard ratio is a very respectable, 0.7. If we go to the next slide, it talks about the next step in immunotherapy. And of course, we shouldn't forget nivolumab, which is the other PD-1 antibody. It was never actually in exactly a trial, as I showed you. Nivolumab was looked at in the trial that I'm about to describe, which was the so-called CheckMate 067 trial, which, for the first time, combined the 2 checkpoints, IPI and NIVO and compared them to the standard IPI alone, but it also included nivolumab alone. And here, the formal statistical comparison was the combo versus IPI alone. And as you'll see, the combo was also better than NIVO alone, but that wasn't what this trial was powered to show. And again, very large, randomized, blinded study, over 900 patients, very well-done trial. And if you look at the bottom line of the trial, which was overall survival, if you go to the next slide, it shows you that, again, with more than 6 years of follow-up, something like 6.5 years of follow-up, there is, a, a plateau for all of the treatments that the combo clearly beats either PD-1 alone in green, that's nivolumab or Ipilimumab in black. And it's pretty obvious that you've now reached a median. But look at the median, which I circled in blue, 50% survival at 6 years. The median survival is 6 years, which is better than colon cancer. It's as good as breast cancer. It's better than prostate cancer. It's way better than lung cancer. So for the first time, you've got a regimen which will prolong survival to the point where your median survival is now at 6 years, which is a phenomenal number. And even though the study was not powered to show a statistically significant difference between the orange curve of the combo IPI/NIVO and NIVO alone in green, it's pretty obvious if I were a patient, I'd want to be on the orange curve and not on the green curve. So consequently, in all academic centers, IPI/NIVO is our go-to regimen if we feel the patients can tolerate the potential for toxicity because the dose-limiting toxicity or I should say, the severe toxicity of grade III and IV with the combo was about 50-some-odd percent. It's much lower at about 20% for NIVO, and it's a little higher at 25% for IPI. But still, the best data we've ever seen in a large randomized Phase III trial in melanoma, 6-year median survival. If you go to the next slide, there was an attempt to take that regimen, which used a relatively high dose of Ipilimumab and a low dose of nivolumab during what we call the induction when you get your first 4 doses of the combination. And then, by the way, you stop the combination, you go on to the maintenance. The current study that I show here, the so-called CheckMate 511 study, was a study where we flipped the doses. So the expression is we give it at a flip dose instead of IPI1 and IPI3 and NIVO1, which is at the bottom, it was NIVO3 and IPI1. So you went from IPI3 to IPI1 and the NIVO went from 1 to 3 mg per kilogram. That was the induction and that was during the first 12 weeks. And then you went to NIVO alone as the maintenance in both arms. So it was a one-to-one study that was a randomized Phase II study of more than 500 -- sorry, more than 300 patients, which was designed to test not the efficacy, but the toxicity. So the primary endpoint of flipping the doses was to see whether you could maintain the effect while reducing the toxicity. And the answer is, absolutely. So if you go to the next slide, it shows both the progression-free on the left and the overall survival on the right. Look at the right-hand curve of survival, whether you have a high dose of IPI or a low dose of IPI, high dose of NIVO, low dose of NIVO, the survival curves are absolutely overlapping. If you look at the orange and the brown curves of survival on the right, I think anyone would agree from across the room, they absolutely overlap. The hazard ratio was 1.03. That's about as close to equalities you can get. Yet if you look at the toxicity, the toxicity went from 48% to 33% at severe grade III and IV, which is a major statistically significant decrease in side effects. So ergo what a lot of investigators now do and what most clinical trials now do is we use the flip dose as our baseline regimen. So in the Phase II studies that I do that are my investigator-sponsored trials, I go with a flip dose plus a third drug. And soon, I'll be going with a flip dose with a third and a fourth drug in my next trial. So this was an important study because it allowed people to feel comfortable reducing the IPI dose in induction with the idea that at 3 years of follow-up, you're going to pretty much maintain the same clinical benefit, but you're going to reduce the toxicity. If you go to the next slide, the next idea was, what do you do about patients with brain metastases. Melanoma is 1 of the 3 histologies that very commonly metastasizes to the brain, the others being breast and lung. And 15 years ago, a melanoma brain metastasis was a dead set. So there had been very little track record of looking at immunotherapy and brain metastases, and this was a straight-up Phase II study of IPI/NIVO at the original doses, IPI3, NIVO1, followed by maintenance at nivolumab every 2 weeks. And this was in patients with brain metastases in 2 cohorts, either you are asymptomatic and you would not had your brain metastases treated or you were symptomatic and you were on steroids and/or you're on steroids. The main data came from Cohort A, which was the majority of our patients with brain mets who pop up on exam or on an MRI with asymptomatic brain metastases. And if you look at the data on the next slide, amazingly, if you look at the intracranial progression-free survival, it plateaus out at 2 years, and that's unbelievable. That means that the patients we'll do well and that half the patients do not have progression of disease at 2 years nor do they have 3 years in the brain. And if you look at the survival curves, it looks just about as good as the original survival curve I just showed you of IPI/NIVO, whether at the regular dose or the flip dose. So if you have asymptomatic brain metastases in a total of 101 patients in a Phase II study, your survival curve on the left is as good as it would look if you had no brain metastases because all the prior studies like CheckMate 067 and CheckMate 511 where IPI/NIVO was tested in melanoma excluded brain metastatic patients. So now for the first time, we can do just as well with a brain met as with no brain mets, which was really very important. And that changed the standard of care. So continue with the next slide. And the most important and most recent Phase III study is not IPI/NIVO, it's not NIVO alone, it's now PD-1 antibody nivolumab with the LAG-3 antibody relatlimab versus NIVO alone, and this is now a 700-patient Phase III study. It matured only very recently. Progression-free survival was the primary endpoint. Why progression-free survival? Because you have other alternative therapies, which you can use after you progress. So PFS is the most important endpoint now. Overall survival and response, of course, are assessed. And if you look at the data on the next slide, it shows that if you look at the primary endpoint of progression-free survival, it's pretty obvious that over time now with more than 24 months of follow-up, the blue curve of NIVO LAG-3 looks a heck of a lot better than NIVO alone. And the progression-free survival is almost doubled from 10.2 to 4.6 months. The response rate is higher, it's something like 42% versus 32%. So you do better with response rate. And at 3 years, the survival is superior by 7 percentage points. So by any assessment, NIVO LAG-3 is better than NIVO alone, and this is definitely now an FDA-approved regimen. It just got approved a couple of weeks ago, and it's going to make its way into the armamentarium probably replacing the use of NIVO or pembro alone. Although the grade III, IV side effects go from [ 10% to 20% ] that's still a relatively low number, and this is now another new element in our armamentarium. If you go to the next slide, looking at the overall survival, as I told you, it's about a 7 percentage point difference. The p-value misses statistical significance because the absolute difference isn't high. But if you look at the way the curves come apart over time with more a follow-up probably by 4 or 5 years, my prediction is that p-value will come under 0.05. And again, because both of these are active regimens, you need to wait a long time to get enough events to occur. And I think that at the time we get enough events, there will be statistical significance. But either way, the appearance of this curve is enough to convince me that NIVO LAG-3 is superior in all respects to NIVO alone. Go to the next slide. Again, if you look at the adverse events, I don't need to harp on this slide. But basically, it's 20% for grade III, IV for the combo, 10% for NIVO alone. It is higher, but is it significantly higher? No. 20% to me is still a well-tolerated regimen. So continue. Next slide. Finally, we can't forget uveal melanoma because, by the way, no immunotherapy really works well in uveal melanoma, which is that rare variance of melanoma that starts in the eye. There are 1,000 cases of uveal melanoma here in the U.S. There are almost 100,000 cases of cutaneous melanoma. And now for the first time, there is a very interesting bispecific molecule called, Tebentafusp, which targets the gp100 melanocytic antigen, which is very commonly overexpressed on uveal melanomas. And even though the response rate was 6%, if you look at the survival which I showed here in this New England Journal article from last year, it's pretty obvious that investigator choice did a lot worse than Tebentafusp, and the FDA approved this -- earlier this year for patients with ocular melanoma in frontline in preference to any other treatment because, frankly, any other treatment doesn't really work in ocular melanoma. So if we go to the next slide, continue. Let's not forget targeted therapy. Again, targeted therapy has a long history in melanoma of so-called BRAF, MEK drugs. I'm only going to show you the most recent trial, which was encorafenib and binimetinib. And if you look at the bottom at progress -- at overall survival, it's pretty obvious that a combination of drugs in the light blue curve beats the reference arm of vemurafenib alone. Interestingly, if you look at the survival of versus of enco, bini versus enco alone, it's not that much better. The enco, bini curve is light blue, and it looks a little higher than the dark blue of encorafenib alone. But the bottom line is at 5 years you've plateaued, but look where the plateau is. The plateau is well below the dotted line at 50%. It's probably about 35%, 40%. That's not what you see with IPI/NIVO, the IPI/NIVO plateau is at 50%. So is immuno-oncology therapy better than targeted therapy? Yes. And if you go to the next slide, that was proven in this study, which is called the DREAMseq study, which just matured and was presented a couple of months ago at the ASCO Plenary. If you look at the sequence of IPI/NIVO, and then if you were failed, you get targeted dabrafenib/trametinib or you get targeted therapy and remember all the patients are BRAF-mutated. And then you get IPI/NIVO if you fail, well, you tell me which arm would you rather be on. I'd rather be on the IPI/NIVO than dab/tram arm in black suggesting that clearly starting out with immuno-oncology drugs is better than starting out with targeted drugs, and you only use the targeted drugs to salvage in patients who are BRAF-mutated who fail IPI/NIVO, which will happen in half the patients. Go to the next slide. Let's not forget about adjuvant therapy, and let's finish up with a couple of trials that clearly show efficacy of adjuvant therapy. There have been a couple of PD-1 trials. This is a trial that I actually was the PI of CheckMate 238, large randomized trial in resected stage IIIb, IIIc and IV melanoma, all high-risk patients who either got for 1-year nivolumab alone or ipilimumab alone, which was approved in the U.S. at the time as a standard adjuvant therapy. You got treated for a year. And then if you look at the most recent updated data on the next slide, it shows that you've reached a plateau at 50%. The median relapse-free survival is 5 years, which is spectacular. It beats ipilimumab alone. So at 5 years, only 39% of the patients have been relapsed. So you have an 11 percentage point increase in relapse-free survival. Ipilimumab does prolong relapse-free survival compared to placebo. So if you were doing a placebo comparison, you'd have a 20 percentage point difference. Similar data have been seen for pembrolizumab, which I don't present because it's less mature. And if you go to the next slide, let's not forget targeted therapy. I told you that targeted therapy as frontline therapy in metastatic disease is inferior to IO therapy. But interestingly, if you look at the results of the COMBI-AD study, which was dabrafenib/trametinib versus placebo in resected stage III melanoma. Go to the next slide. If you look at the plateau at 5 years, it's the same plateau as you saw with nivolumab or with pembrolizumab. And amazingly, a drug which is -- or a regimen which is inferior to IO drugs in metastatic disease looks just about the same in resected melanoma as an adjuvant therapy, so go figure. And again, this is a placebo trial. So the placebo control is down to 36%, that's a 16 percentage point absolute difference. That's a big deal. That's very, very impressive. So now we have dab/tram, we have pembro, we have NIVO, all are effective adjuvant therapies in melanoma. And lastly, if you go to the next slide, we have looked at the use of adjuvant IO therapy in resected stage II disease, and this is the recently matured CheckMate -- or KEYNOTE-716 study. This was pembrolizumab versus placebo in resected stage IIb and IIc melanoma, whose outcome is better than those who have stage III and IV resected disease. Straight up, randomized study, 900-plus patients either got pembro or placebo for a year, and then you got followed. And if you look at the next slide, it shows the data. Now not as distinctly different as with stage III, IV disease, but clearly, the relapse-free survival is superior for pembrolizumab alone versus placebo. The hazard ratio is respectable. It's 0.61. This regimen just got approved in the U.S. It will probably get approved in the EU. And now we're treating patients with resected stage IIb, IIc melanoma with adjuvant pembrolizumab just approved by the FDA. Go to the next slide. I'll end up with reminding you that a lot of the action in melanoma is now a neoadjuvant therapy in patients with palpable stage III disease, which, frankly, is not that common. Only about 15% of the patients we see who have stage III melanoma have palpable disease. But if you do, you can do a neoadjuvant scheme where you do a pretreatment biopsy of your palpable lesion. You give drugs for 6 to 8 weeks, you then operate, and you can then do all kinds of analyses of the resected specimen to tell you what correlates with outcome. And again, if you go to the next slide, it just shows you a snapshot of a bunch of trials, and it shows if you give neoadjuvant immuno-oncology drugs, you have a very high chance of getting a pathologic complete response that's PCR. And that means that all of the tumors dead, even though the tumor radiologically may not go away when you take it out of the patient, it's dead. And that happens in about half the patients who get IPI/NIVO for 2 cycles. And if you look at that blue line on the left of recurrence-free survival or the blue line of overall survival on the right, those patients do fantastically. If you look, however, at the relapse survival curve in yellow, that's if you don't get a pathologic CR, that's where all the relapses come. So what this means is if you give neoadjuvant therapy and you get a pathologic CR, which happens half the time, those are the patients who are going to do well. If you don't get a pathologic CR, almost all those patients eventually going to relapse. There now is a randomized study of neoadjuvant surgery adjuvant versus adjuvant therapy alone, we'll find out whether giving neoadjuvant therapy is better than just giving adjuvant therapy after surgery. So if you go to the conclusions, just let me remind you that PD-1 blockade alone with pembro or NIVO gives a 44% response rate and a median survival of about 3 years. The long-term follow-up, however, of IPI/NIVO and also IPI/pembro in smaller studies shows excellent 5- to 6-year survival with a 6-year median survival. But the rate of grade III, IV side effects, which are the severe ones, it's 55%, that's really high. LAG-3 NIVO in frontline melanoma has better progression-free, overall survival and response rate. But with a 19% or 20% rate of grade III/IV immune-related adverse events, which is higher than the 10% you see with PD-1 alone. There's excellent activity of IPI/NIVO which applies for IPI/pembro and IPI/NIVO, which is superior to NIVO alone in patients with brain metastases. Adjuvant treatment with nivolumab, pembrolizumab, dabrafenib, trametinib has a hazard ratio for relapse-free survival of about 0.5 to 0.6 versus no therapy. That's fantastic. That means you cut in half the risk of relapse. Neoadjuvant therapy with IPI/NIVO is the best neoadjuvant therapy we have, and it probably picks out the patients with long relapse-free survival and the sequence of IO drugs then salvage with targeted drugs clearly superior with the converse of targeted therapy than IO. So we've done very well with melanoma, But trust me, half the patients that get treated are still going to die of melanoma. We've got a long way to go and a significant unmet need. Thanks for your attention.

All right. I'm going to jump right in there, talking about personalized cancer vaccines. Thanks very much for the opportunity to present this today. Next slide, please. Here are my disclosures. Next slide. So when you compare cancer vaccines to vaccines against infectious pathogens like SARS-CoV-2 vaccines, cancer vaccines are typically done under therapeutic setting versus the vaccines against infectious pathogens are usually prophylactic given to healthy individuals. Cancer vaccines really can generate new antigen-specific T cell responses against tumor cells, but they can also amplify existing responses and increase the breadth and diversity of the tumor specific T cell repertoire. Please go to the next slide. Neoantigens are encoded by tumor mutations and are attractive antigenic targets for vaccines because they are novel to the immune system, and they're also highly specific to the tumor. This really has led to a paradigm shift away from neoantigens towards neoantigens as targets for cancer vaccines, but also for other modalities like adoptive cell therapy. This talk really focused on the vaccine approach. Because neoantigens are almost exclusively unique to individual tumors, target then requires a personalized approach. Next slide, please. The customized process that's necessary to manufacture a personalized neoantigen vaccine and principle is shown here. For the target selection, each patient's tumor and normal tissues are sequenced, limitations are then identified and epitopes are selected based on several variables, including MHC binding prediction models, RNA expression, but also similarity to self sometimes similarity to foreign pathogens, formality and other variables. And there is now many computational algorithms and machine-learning tools, as we will hear later on, that have been developed to identify mutations in sequence data in order to best prioritize those that are thought to be the most immunogenic for neoantigen targeted approach. As you all know from the SARS-CoV-2 field, there are many different platforms for cancer vaccines. Next slide, please. There are some high-level considerations for the development of personalized neoantigen vaccines. Various factors should be considered for the design. First and foremost, the time that's required to generate the personalized vaccine, which usually is on the order of several weeks to months is a critical factor, particularly in the metastatic disease setting. There are 2 variables really that account for that. One is like it's the process of identifying the best neoantigens. The other one is the time to manufacture those individual vaccines. And then for the design of these trials, additional variables include the route of administration for the vaccine, the dosing regimen, and of course, any sequencing approaches or combinatorial agents that are being used. Next slide, please. To test the concept of vaccinating against neoantigens in cancer, at Dana Farber, 10 years ago, we conducted the first-in-man trial. And we did this in patients with high-risk melanoma using personalized long peptide vaccines. These vaccines were formulated into 4 distinct pools and they were with admixed Poly IC LC agent as an adjuvant as for immune stimulation. And the vaccines were given subcutaneously on a prime boost regimen, as you can see on the lower part here of the slide. Next slide, please. In this study, we showed really robust immune responses against the multitude -- or the majority of the vaccinating epitopes. We presented the first 6 patients and showed really like striking responses that we're seeing ex vivo and interferon-gamma ELISPOT assays. When we deconvoluted these responses, we saw a predominance of CD4 T cell responses. And we saw ex vivo responses with the CD4, but not with CD8. Next slide, please. We also were able to show that these vaccine-induced T cell responses were specific for the mutant versus the wild-type epitopes. In 3 out of the 6 patients, we showed reactivity against autologous tumors, which is important. And then we had 2 patients that actually had a recurrence while they were vaccinated, but then were treated with pembrolizumab and both of them had a complete responses that were at a really rapid onset, which was -- which we interpreted as potentially an early signal for activity simply because those CRs are relatively rare even in melanoma and certainly typically not a rapid onset of PD-1 inhibition alone. We also saw expansion of immune responses when we use PD-1 inhibition, which we interpreted as a good rationale for a combination with PD-1 inhibition. Next slide, please. Another melanoma trial testing a personalized vaccine using an RNA-based approach done by [ Brigham-Tech ] a similar level of vaccine specific T cell responses were seen and there was also this predominance of CD4 over CD8 responses that was observed. Next slide. Follow-up analysis of the melanoma trial done at Dana Farber, we saw functional exchanges and TCR diversification over time when we analyze single-cell transcriptomes and TCR clonotypes of tetramer-sorted vaccine-specific T cells. Furthermore, in bulk T cells, we observed persistence of vaccine -- immune-induced -- vaccine-induced immune responses over several years, actually up to almost 5 years out. And we also saw a spreading of vaccine-induced responses to nonvaccinating epitopes, which I'm going to describe in a little bit more detail in another trial. This could potentially indicate tumor cell killing mediated by the vaccines. Next slide, please. And here is the design of this trial, there was a follow-up study and that was sponsored by Neon Therapeutics, which is now BioNTech US. In this trial, patients with 3 different -- with 3 different cancers, melanoma, bladder cancer and lung cancer were enrolled. The patients were naive for PD-1 inhibitions. Most of the patients were treated on the first line, and they received nivolumab while the vaccines were manufactured for the first 12 weeks. Started at week 12, the vaccines were given on a prime-boost schedule, just like in the first trial at Dana Farber and the nivolumab was continued. The objective of the trials were safety, efficacy, and a pretty deep dive into the immune responses as well. Next slide. As you can see here, the immune responses were measured at 3 different time points. This is really in line with the design of the trial where, as I said, that the patients started with PD-1 inhibition. So we obtained luciferases samples and tumor samples prior to the PD-1 inhibition and then before the vaccine and after vaccines, so 3 serial time points were available for those immune assessments. Next slide, please. Here, you can see what is arguably one of the biggest challenges for a personalized approach, which is -- by which I mean the drop off that it can get when you have metastatic cancer patients that have to, in some way, wait for the vaccines and so this time delay. And you can see that there is some drop off from the time of enrollment to the time of vaccination and then for the patient that actually completed all the vaccines. This drop off was relatively modest in melanoma patients, but was higher in the lung cancer and bladder cancer cohorts. Next slide, please. Nevertheless, similar to the first study, we also did a comprehensive analysis using interferon-gamma ELISPOT and saw responses against, again, roughly 50% of the immunizing epitopes. Again, a predominance of CD4, where 40% of the vaccinating epitopes induced CD4 responses and about 20% to 30% induced CD8 responses. These responses were largely de novos or not seen prior to the vaccination. Examples are shown for the individual cohort when we deconvoluted the individual vaccinating epitopes. Next slide. Looking at waterfall plots in the 3 tumor cohorts, response rates were roughly in line with historical controls, albeit a little bit higher if you just look at, for example, the response rate of 60% when typically response is more 40% in melanoma. Nevertheless this is obviously a cross-trial comparison, and there's probably some patient selection. Certainly, I'd say, encouraging, but not definitive in terms of the clinical signal. We did see deepening of the responses over time, which we attribute at least in part to the vaccines, although admittedly, although also PD-1 inhibition sometimes leads to delayed or deepening of the responses by itself. Next slide. We are also determining whether neoantigen-specific CD4 and CD8 T cells in the peripheral blood had a cytotoxic phenotype by looking at a molecule called CD107a, which is a market for T cell granulation. And we saw -- we had 21 patients available for this assessment and saw 60% of the vaccinating epitopes had induced T cell responses where there was degranulation seen, suggesting cytotoxicity of the vaccines. Next slide, please. I mentioned the phenomenon of epitope spreading a little bit before. This is like a way to potentially delineate vaccine-induced versus PD-1-induced activity. The idea is that epitope spreading in a setting of a tumor response is really a phenomenon where tumor antigens are released due to tumor destruction and thereby triggering additional immune responses. Next slide, please. And so on this study when we looked at the -- when we discovered the neoantigen, there was essentially typically more mutations or more neoantigens that were, in theory, candidates for targeting and only a part of them made it into the vaccine. So this provided an opportunity to look at responses against vaccinating versus nonvaccinating neoantigens. Next slide, please. And this is what was done here where you can see responses to nonvaccinating epitopes that were detected the post vaccination, but were not evident prior to the PD-1 inhibition or prior to the vaccination. And those responses were seen in 3 of 18 patients who had prolonged progression-free survival, but only in 2 of the 7 patients without prolonged progression-free survival. Go in the next slide. So this association of epitopes spreading and prolonged progression-free survival is shown here for the aggregate data set in 22 patients with the 3 different tumor types. And so that, that could be a surrogate again for vaccine-induced tumor cell killing. Next slide. As I mentioned on this trial, we had serial tumor samples available for interrogation. And so we're able to look at pathologic responses in relation to treatment with PD-1 inhibition and vaccines. I will point out that those pathological responses are different from what Dr. Weber just presented, where those were core biopsies where we were able to look at tumor content, not surgical samples of an entire tumor. Nevertheless, when we assess tumor samples in 14 patients with metastatic melanoma, 5 of the patients shown here in blue had complete pathologic responses after 12 weeks of nivolumab prior to the vaccination. So that's great for those patients, but not particularly helpful to actually look at vaccine-induced tumor activity. Go to the next slide, please. Over 9 of the 14 patients shown here in green did not have complete responses prior to the vaccination. However, all of them achieved pathologic CRs after the vaccination. again with a stipulation that those were core biopsies not surgical samples. And then please go to the next slide. Interestingly, these pathologic responses were also just like with the epitope spreading associated with improved progression-free survival. Please go to the next slide. There's many clinical trials that are ongoing mostly really in industry. However, some academic centers are also conducting personalized vaccine trials. These trials are using different platforms, as I mentioned. Here just shown a sample of about 10 trials where peptide approaches are shown in yellow. RNA-based approaches are shown in red, DNA in blue and viral based in green. Go to the next slide, please. Recognizing that those are too many efforts and not even all of them are actually shown on this slide to present here, I decided to discuss 2 programs in some detail in the following slides. So one is a trial using an RNA-based prime-boost approach taken by a company based in Cambridge and in the Bay Area called [indiscernible]. This is indicated here in the red frame. Next slide, please. The vaccine delivery approach developed by [indiscernible] consists of a replication-defective adenovirus for priming and self-amplifying RNA for the boost. This vaccine is given along with -- or in some of the trials given along with subcutaneously administered CTLA-4 inhibition and showed quite robust CD8 responses in nonhuman primates, as shown here on the right side of the slide. Next slide, please. On the first-in-man trial called Granite, patients with gastroesophageal cancer and then MSI-stable colorectal cancer or non-small cell lung cancer received escalating dose of vaccines along with a CTLA-4 inhibition, as I mentioned. 14 patients were enrolled in the dose escalation in the Phase I part of the trial and 12 patients have so far been treated in 3 different expansion cohorts in the Phase II part of the trial. Next slide, please. Overall, robust ex vivo specific T cell responses were seen in the majority of the patients. And in the microsatellite stable colorectal cancer cohort, a correlation between decrease of circulating tumor DNA and prolonged overall survival was seen and was interpreted as an early signal for full efficacy of the vaccines. Next slide, please. Next, I'm going to give a quick overview of another trial called the iNeST trial, which was conducted by a collaboration between BioNTech and Genentech/Roche. And this platform is based on an RNA vaccine, which was really actually generated through the first study that I showed you that was done in melanoma and provided a platform for the SARS-CoV-2 vaccines that Pfizer and BioNTech now is world famous for, so to speak. Next slide, please. So this drug, which I'm not going to spell out here, there are a lot of numbers is a systemically administered RNA-lipoplex vaccine that's really designed to stimulate T cell responses against neoantigen. The vaccine is given intravenously and was previously in a publication shown to be delivered to dendritic cells in the spleen. The target number of neoantigens is also 20. And because it's RNA, it actually has intrinsic adjuvant properties as a TLR7/8 agonist. Next slide, please. In the first-in-man trial called iNeST, as I mentioned, patients with advanced or recurrent solid tumors received escalating doses of the vaccine in the Phase Ia part, followed by dose escalation of combined therapy with vaccines and the PD-L1 inhibitor, atezolizumab. The vaccines were administered every week for 7 weeks and subsequently, every 24 weeks versus atezolizumab was given every 3 weeks. The primary objective was safety in Phase I and RP2D assessment in the Phase II. There's like a bunch of expansion cohorts, as shown on the right side of the slide. Next slide, please. Overall, ex vivo T cell responses specific for at least 1 vaccinating neoantigens detected using ELISpot or MHC-multimers were seen in 3/4 of the 63 patients who were evaluated in this study. So showing again, pretty consistent immune reactivity against the vaccines. Next slide, please. And interestingly, there were some patients that had really like very high magnitude CD8 response up to more than 5% of the peripheral blood, which is really quite striking. These T cells show an effect on memory phenotype, which is what would be expected in a sort of -- encouraging in terms of the right T cell population being induced by this vaccine. This is a patient with triple-negative breast cancer. Next slide. And this is actually so immune data on the right side, again, a breast cancer patient with a robust immune response. However, so this is the dose escalation part of the trial. So not every patient got actually a sufficient vaccine dosing. But as you can see, there were 2 responses. But overall, the efficacy was rather modest. As you can see that most of the patients actually did not have responses. But again, I will say this is like the Phase I part of the study that show. Next slide. And so to summarize what we can say about personalized vaccines, the immune response are certainly across the board quite robust, but they don't really quite yet consistently translate into clinical responses. So I've shown you the immunogenicity data. We have data -- really long-term responses that last for several years. I showed you the epitope spreading, the pathologic responses after the vaccines. Nevertheless, there's certainly room for more work in this space, there is, as I alluded to, many approaches being taken, many vaccine platforms. Next slide, please. I will point out sort of the main challenges that the field really has, I'd say, as like really time and cost, right? Obviously, a personalized approach is costly. I will say because this is -- the time it takes to make the vaccine is largely sort of technological in some way, so there isn't really a time that it takes to grow sales or so far as, for example, with a cell-based approach. So these time lines can really be compressed and have already been compressed by some of the companies that are in this space. I would say one can certainly improve on the robust -- even on the robustness of the immune responses, particularly the CD8 responses that we have shown in our trials where we didn't really have ex vivo or CD8 responses. CD8 responses were not consistently seen in other trials. So that's certainly something that could be worked on. But obviously, there's lots of things that can be done. One can use different immunoadjuvants that can work on the dosing schedules. One can certainly -- and the really important point is like what is actually the best immunogenic neoantigen. That is something that is really a critical question in the field like how do you best in silico determine what the best neoantigen to vaccinate against this. Next slide, please. And so I'm just like a couple of sort of suggestions and thoughts that should be done, can be done like in the peptide vaccine approach world as like opportunities to make faster manufacturing of the vaccines just simply better peptides. But obviously nucleic acid approaches can be taken and are being taken and then one could think of like ex vivo delivery of -- with different immune cell types, so sort of hybrids, cell-based vaccine approaches and then also in vivo delivery of immune cell populations. And then I already alluded to combinatorial approaches and sequencing. There are certainly even PD-1 inhibitor is the most logical partner, there's now data coming out that maybe starting out with PD-1 inhibition is actually not ideal because it can negatively impact the memory T cell population. So a strategy to initially use the vaccination first and then follow up with PD-1 inhibition may actually be better. And so I'm just going to stop here with this last slide where I just sort of point out that the sort of the engine source is really this neoantigen discovery part. And then the vaccine delivery approach are the 2 critical things that I feel can be and should be addressed in this field. Where really what is the right neoantigen and how deliver it best that these 2 aspects really working for injunction. And I'll stop here but just acknowledging some of the players at Dana Farber, and I'm looking forward to the discussion. Thank you.

So next slide, I suppose, thanks a lot to both our presenters for an excellent update, much appreciate it. So this table is a quick glance at our development pipeline. We will be focusing on oncology marked by the green bars, and I'd like to direct your attention towards the EVX-01, 02 and 03 tracks, and they include both peptide and DNA-based [Audio Gap] therapies. Well, as you may see, not the topic of today. I'd like to underscore that the foundation of our technology is called PIONEER. This is an AI-based tool that allows for accurate prediction and ranking of personalized immunogenic neoepitopes. Next slide, please. This should show the -- our first clinical trial using EVX-01. This is -- incorporates metastatic melanoma patients being immunized using a peptide-based vaccine at increasing dose levels. Next slide, please. And this is just briefly outlining our findings, which were that all primary and secondary endpoints were met. EVX-01, our peptide-based vaccine, appears to be well tolerated with only Grade 1 and 2 adverse events have served. Also, the overall response rate was at 67% and a complete response rate of 22% in combination with anti-PD-1 therapy, which compares favorably to anti-PD-1 monotherapy. Furthermore, 3 of the patients with a stable disease for 8 months or more on anti-PD-1 therapy transformed into 2 complete responses and 1 partial response after receiving EVX-01 therapy. Also, we saw a broad T cell activation in all of the patients with a large fraction of the PIONEER identified new epitopes, including a de novo response -- inducing -- sorry, inducing a de novo response. We also saw a correlation between T cells being activated by the PIONEER-identified neoepitopes and the clinical response. Furthermore, we saw a correlation between the EVX-01 activated T cells and the antitumor effect. And finally, a recommended dose for Phase II was established through this trial. So the next slide, please, which should show you that in collaboration with Merck, we are now preparing a follow-up global Phase II trial based on the prior trial that I just showed you. So this is designed to ensure a fast, clear and clinically meaningful outcome with decision points for potential partnership. It is a single-arm trial evaluating the efficacy and safety of EVX-01 in adults with unresectable and metastatic melanoma. It will be conducted in Australia, across Europe and in the U.S. And in terms of milestones, we filed -- we had a regulatory filing towards the end of 2021 and first patient, first visit is expected this side of summer. So next slide, please. And again, this is just a rapid walk-through of our clinical program. We're now moving into our DNA-based vaccine section. We've shown promising results in a preclinical trial -- in preclinical trials using both EVX-02 and EVX-03. The latter being our optimized construct incorporating our proprietary APC-targeting unit, but both modalities accepted a highly protected effect in these preclinical trials and being seemingly dose-dependent. Next slide, please. So this depicts our -- an outline of our Phase I trial using the EVX-02 DNA-based vaccine. It is currently being conducted at 6 centers across Australia. Enrollment was completed recently, and it involves adjuvant melanoma patients being treated with nivolumab. We're looking into safety, tolerability, feasibility as well as some -- various clinical observations. And we take some -- we're happy to see also that we were able to provide a needle-to-needle turnaround time of approximately 50 days on average. Next slide, please. So based on some of these encouraging preliminary findings in the first couple of patients in the DNA-based program, we are now awaiting more robust readout from all of the patients this year -- later this year. But I'd like to draw your attention to the table at the bottom, in which you'll see readouts from 2 individual patients. And you'll see that we were able to detect reactive neoepitopes using both ex vivo and IVS ELISPOT, and perhaps equally or maybe even more importantly, we were able to see the intracellular cytokine staining and thereby identifying CD4 and CD8 positive T cells. The majority of these new epitopes that were incorporated. I might mention that for this particular trial, we incorporated 13 new epitopes. In the previous trial using the peptide-based vaccine, we're incorporating up to 10 different neoepitopes. And we're coming to the end of my little presentation, the finish line. So next slide, please. That's the final slide that I have before we engage in a Q&A. So based on these recent findings and the ongoing trial using the DNA-based program, we are looking to advance this DNA-based vaccine program further into the clinic by opening up an additional Phase II trial later this year. So we're currently engaging externally, while also fine-tuning the design. And this will be based on our optimized EVX-03 platform. So with this, I thank you for your attention, and I will actually hand it over to you, Tara.

Great. Thank you, Erik. [Operator Instructions] So our first question comes from [ Zeike Heff ] from Oppenheimer.

Can you hear me now?

Yes, we can.

Really 2 for -- one for Dr. Weber and one for Dr. Ott. As for Dr. Weber, when -- you mentioned that like the clinic melanoma has been evolving and there are different combinations. So when we think about conducting a like controlled study for either from melanoma or adjuvant melanoma treatment, what kind of like comparator, like which treatment regimen would you think as a proper comparator arm? And also, what kind of evidence would you think as positive for patient adoption?

Well, I think right now, the -- in stage III resected melanoma, the comparator would be NIVO or pembro, you pick them. There is a very promising large randomized study of NIVO LAG-3 versus NIVO going on. That's not going to finish for at least a year. It's not going to mature for at least 3 years. So if I started a study today, NIVO or pembro would be my control arm. If I started it 3 years from now, I wouldn't bet against that combo being positive. So I would say it could be NIVO LAG-3, that would be perfectly reasonable. Or if I did it in Stage IIb IIc resected patients, which would be a great place to do it, there are no stage IIb, IIC randomized trials going on now. It would be either -- would be pembro, which is the one approved drug that would be my control arm. And personally, I do a randomized Phase II. I'd get it done quickly. I treat 120 patients with a 2:1 randomization vaccine pembro versus pembro alone. And if it looked promising by the first year, I'd go straight to a Phase III study. But that's me, I'm a conservative trialist. I don't believe in going from Phase I studies to Phase III studies without stopping in between.

That makes sense. And for Dr. Ott, like with the challenges of your confirmation for neoantigen development, of prior experience, which patient population do you think will have great benefits with the neoantigen cancer vaccine Is that frontline patient or new adjuvant or adjuvant regimen?

Well, -- the short answer is I don't know, right? So there's a reason to test this in the metastatic setting in order to get like a signal. Obviously, there's like -- there's always this thought that an immune -- that a vaccine could work best in an adjuvant setting, right, but much harder to really get a signal of single-agent activity of the vaccine itself. The best way to show activity is always a feel in the PD-1 resistant setting, right? So we picked up a response in patients who are resistant PD-1, I think that would provide the clearest signal for efficacy going forward. That hasn't really happened yet with -- at least from what I'm aware across the field where it is. Although there have been like single key reports of patients actually had responses with personalized vaccine plus PD-1 inhibition -- actually about PD-1 resistance. But I don't think -- so I will say that I feel one of the attractions of the approach is to actually not like exclude certain populations. Because in general -- in principle, this approach can work for any patient population. It's just about finding the right and the best neoantigen. And I would say even in a cancer that has like lower mutation rate, that can be feasible, although it will be more challenging, right? So it's certainly much easier to pick or to find enough neoantigens in the melanoma compared to colorectal cancer. But that doesn't mean -- we don't really know how many you actually need, right? So if you need just like 1 or 2, and those are sufficient, I think it's still possible. And there's actually a proof of principle and the adoptive cell therapy field has already been shown, where -- there's like a report from cholangio cancer patients who had like a nice kind of a response with an adoptive cell therapy approach.

Our next question comes from Ahu Demir from Ladenburg.

This is a great event. Thanks for the KOLs putting the landscape really well. We appreciate that. My first question would be, I think, to Dr. Patrick Ott. You have mentioned the immunogenicity data not translating into the clinic. I am curious if there is anything that we can pay attention -- particular attention to early clinical signals, would it be perhaps the pathological complete responders? Is there anything that we know that actually might give us better responses and we learned the lessons from the previous trials, perhaps?

Yes. So I didn't -- when I said it doesn't translate, I meant more like directly, right? So what we do see is like really consistent robust responses, which is like novel for the cancer vaccine field like before most of the trials, most of the vaccines were seen in like subsets of patient populations and were sort of modest at best. Now we see like really consistent ex vivo responses, which is very striking. But we don't really -- we don't know whether a patient has like a response like is actually a responder. I think there -- so I tried in my presentation to show things that we can look at, like the put up spreading, like these path responses. It's a little -- I would still not say like those don't -- still don't directly translate into RECIST responses. I don't think we have really figured out kind of like what the right surrogate is. I personally think that -- so our first -- and there's obviously [ enough ] 2, but sort of those 2 patients that are presented in our first trial, where we basically -- those were the 2 patients who had metastatic disease, they both had like recurrent tumors. They have received the vaccines and then they got the PD-1. They both had the responses. Despite the low numbers, I thought that was like a pretty striking example of like how this could potentially turn. We haven't really -- it's -- the trick is actually to replicate that in the clinic consistently because it's like -- it's difficult to design a study like that. It's difficult even in an adjuvant study to have a personalized vaccine that takes like weeks or months to generate while a patient, for example, does not get PD-1 inhibition. Because I think another thought that I feel we're definitely better to vaccinate before the patients get the checkpoint inhibition. So the solution to do that is really to compress the time line, right? So if you could do it a vaccine or you could generate the vaccines quickly, that may be feasible. But it's just back to you like what you were asking. I don't think we really have like a good surrogate where we basically just now, if there's a epitope spreading, those patients will then also have RECIST responses. I mean I wish we had, but we didn't detect that, for example, and then do a study.

I would like to ask a question to Dr. Weber. Considering your expertise, knowing the field really well, I'm curious what you think about vaccines? And we have other, of course, combination strategies, oncolytic viruses, TLRs and many, many other approaches. Just curious what you think about the vaccine approaches? And where is the biggest unmet need that the vaccines could address?

I think the biggest unmet need that the vaccines could address would be either in the adjuvant mode as Evaxion is testing or in a scenario where a patient has achieved a partial response or stable disease and then you vaccinate them and then you might even add other immunomodulatory agents. As you probably know, patients who achieve a stable disease, 70% or 80% of them are going to relapse within the first couple of years in melanoma, 30% to 40% of the partial responders will progress. So there is a significant number of those patients that will give you a relatively rapid clinical endpoint. In the adjuvant mode, if you choose a Stage IIIc or 4 resected patients, I can guarantee the rate of relapse at 1 year will be very significant, probably 40% of them are going to relapse in the first year. So you got a pretty rapid turnaround. An endpoint that within a reasonable amount of time will yield useful information. So I think that's where the unmet need is. I think the best utility of cancer vaccines, at least in melanoma, is going to be in the adjuvant and in the mode of treating the stable and partial responders. Treating a complete responder, you won't need to do because 85% of them are going to stay in remission if they achieve a complete response. It's the ones who don't get a CR, those are the ones you have to deal with. So that's what I see as the fertile field for vaccine treatments. In addition, I also like the idea of vaccinating someone and then look for easing them to generate T cells that are neoantigen-specific since you've enriched peripheral blood to a very significant degree for the neoantigen cells in most cases with the vaccine. You can then luciferase the patient, expand the cells using peptide pulsed dendritic cells or peptide pulsed artificial APCs. And then add the adoptive cell transfer to the vaccine strategy, you then give the adoptive cell transfer and then you vaccinate them after the adoptive cell transfer, which should, again, prolong the persistence of those cells and amplify their proliferation.

I have one last question. I'm curious at the AACR we have seen data from Genocea Biosciences. I'm curious if any of you might have a comment given the first reaction, I think?

I haven't seen the data yet. No earlier data there, but I haven't seen the AACR presentation, yes.

Well, could you summarize the data? It sounds like neither of us are in the loop.

I don't think I can summarize the data. I have seen the poster. But yes, I don't think I can now.

Our next question comes from Thomas Flaten from Lake Street Capital Markets.

Dr. Weber, I just wanted to make sure I had a better handle on your thought on the PD-1/LAG-3 combination. Given the IPI/NIVO side effect profile, do you think the PD-1 LAG-3 will quickly overtake being first line utilization? Or can you just walk us through, again, your position on that?

Sure. I think that NIVO LAG-3 will replace the use of pembro or NIVO alone in patients who are older, patients with preexisting autoimmune issues and anyone who is not in a position where they could tolerate significant therapy. In academic centers, we will continue to use IPI/NIVO as our frontline regimen of choice if the patients can tolerate the potential of the high toxicity, which even at a flip dose, is still at 33% Grade 3, 4 immune-related adverse events. But anyone who we would, let's say, an 85-year-old who's a little fragile that we would have treated with pembro alone, I'd give him NIVO LAG-3. I think it's a better regimen, better chance to benefit, and the side effect profile will still be very favorable.

And do you -- are you able to profile those patients beyond the kind of obvious who are going to have significant tox issues?

I wish I could. If I could predict who's going to have toxicity, I wouldn't be here. I'd be sitting in the Pacific Island shipping on a Mai Tai.

Dr. Ott, I was curious to get your take on some of the work that's going on around using their described various ways, but hotspot mutations, TP53, EGFR, KRAS, et cetera, as an alternative vaccination strategy maybe circumventing the time lag involved in developing a personalized vaccine. And I realize it's complicated by HLA-typing, et cetera. But I'm just curious to get your general thoughts on that before I have a follow-up.

Well, I wouldn't say it's not complicated. It's basically prohibited by HLA, right? So you have like -- you said substantially lower the patients that are eligible for such a trial. That doesn't mean like there could be inventive ways to like create peptides, for example, that have like less restriction to HLA and there's actually efforts ongoing. And I'm kind of way off -- where you may actually be able to generate a vaccine against recurrent mutations such as KRAS or P53 that has -- that is not quite as restricted. And so therefore, not as limited. I think it's obviously like that would be great. I just think that so -- in our efforts when we look for the best neoepitopes, and obviously, there's other factors that go in there when you look at the immunogenicity, right, just HLA binding is not going to cut it necessarily, right? You want to know whether -- how well that mutation is actually expressed. And similar, as I alluded to, a similarity to like self and pathogens and all kinds of -- the likelihood that's going to be processed by antigen presenting cells or cancer cells. So a lot of these things would not necessarily fit the bill in such a vaccine for each patient. So I think there's certainly -- there's definitely these challenges, but I would be -- there's no question that would be better to have like an off-the-shelf product. But I just don't think one could really capitalize on the full potential of neoantigens that way.

All right. And then just one final one for me. Given the -- you obviously have a view of the totality of the data or much more so than I do. If you look at the -- or are the differential responses by CD4 versus CD8 activation? And do you have a sense of if it's going to lean one way or the other going forward? Or do you think they can both be potent or deliver potent responses regardless of CD4 CD8?

Well, I mean, in a classical sense, like speak to any immunologist, they will tell you that you need CD8 response -- CD8 T cells to like really induce effective T cell. I'm not questioning that. I will say that CD4 -- including specific CD4 T cells are probably important to actually have cytolytic capabilities themselves. So it becomes increasingly clear that CD4 T cells also play a role. But I mean it would be -- there's no question that would be nice to have like a better CD8 or improved CD8 responses. I do think there's a good explanation for why we see these predominance of CD4 T cells. That's basically a function of the vaccine. It's just like easier to generate CD4 T cell response than CD8. And it might also be to some extent, a function of like their readout. But nevertheless, I think they are important. I think both are important. We need to like work on sort of make the CD8 T cells better, and don't -- not forget about the CD4, but I would not say it's like the holy grail that only CD8 cells will do the job.

This concludes the verbal portion of our Q&A session. I'll now turn it over to Corey Davis from LifeSci Advisors to read the questions that came over the webcast.

Great. Thanks, everyone. We actually covered a lot of the questions that have come in already. And in the interest of time, I think we just have probably 2 questions that I'll ask. The first one for Dr. Ott. Could you elaborate on the importance of correctly identifying not just any neoepitope, but ranking these neoepitopes to ensure that the most potent or immunogenic ones are incorporated into any vaccine?

Well, I mean, that's the intent of all neoantigen discovery approaches is basically to prioritize which means basically come up with a list where -- like which one is the best and then go down the list and then basically incorporate them into the vaccine. At least that's what's really being done with those with in silico approaches is really to come up with the list, but also have ranking of -- but by nature, I mean, those algorithms will almost really by design come up with like a prioritization because it's like -- there's like basically a number of variables, and you can compute those variables and it will give you like #1, #2, #3, #4. That's an easy thing to do. I'm not saying -- but what's not easy is actually to accurately predict and actually validate the best neoepitopes, right? That is really like what we've still don't now. There's been a lot of effort. They have been consortia that have been trying to get at this, the one called TESLA like an effort by Parker Institute. I would say like we don't -- we still, despite these efforts, don't have like a -- like 1 platform or 1 tool where we can really reliably pick the best ones. And I think it's an issue in the field. But it certainly also tells us like, well, there's a lot more work to be done, and there's a lot of opportunity there.

Great. For Dr. Weber, you did a great job of identifying still the unmet medical need despite the advances in the field with half the patients still dying out the past 7 years. But you also mentioned that you could see using 3 or even 4 drugs in different settings. Could you elaborate on that statement? Is there anything obvious where a triplet could eventually become the standard of care in melanoma?

The reason why I said that is based on a clinical trial that I'm involved in with several colleagues at Cedars-Sinai in Los Angeles. Patrick's own institution at Dana Farber. So he knows the trial and Mass General, where we give IPI/NIVO at what we call flip doses with tocilizumab, which is the IL-6 receptor blocking antibody. And it's based on a whole series of -- a whole set of data suggesting that IL-6 levels are associated with the poor outcome with immunotherapy and with melanoma in general. And the fact that if you block IL-6, you can reverse steroid-refractory immune-related adverse events. And indeed, in the first, we've treated probably 58 or 59 patients. It appears like we can reduce toxicity. The clinical efficacy is unimpaired. So it looks pretty promising. And our next study is the reason why I mentioned 4 drugs. We're going to go IPI, NIVO, relatlimab, sarilumab. So that's my next study. That's 4 drugs.

Excellent. Great. Next question is probably for Erik, just more corporate in nature. But also you guys can comment as well, but could you describe the patient population that was treated in the EVX-01 Phase I study and compare that to what you're planning for Phase 2, differences or similarities in the inclusion criteria?

Yes. Absolutely. Thanks, Corey, and thank you for allowing me to just provide a little bit of a granularity, yes, I appreciate it. It was a quick one through. So in the EVX-01 Phase I trial, we looked at -- or we included patients that were either treatment naive. And also, there's a different subset of patients that had been treated with checkpoint inhibitors for at least 4 months and that we're exhibiting no progression, yet no real resolution or -- so no progression or no complete response. So those were 2 different subsets. For the -- and that we're obviously in metastatic melanoma for the Phase II program. We're looking also into incorporating patients with metastatic melanoma but with a much more uniform population, we will be strictly mimicking the inclusion and exclusion criteria adopted in the KEYNOTE-006 trial, that's the pembrolizumab trial. And by that, we hope to demonstrate a clinically meaningful difference compared to pembro alone.

Maybe we'll do one more question. This is probably for Dr. Ott, given your comments about the manufacturing time. What do you think is the ideal amount of time for manufacturing for neoepitope-based therapies to be able to retain the patients that enroll in the clinical studies?

One day. I know. I mean I can only say like the faster the better, right? But I will repeat what I said, I do feel there's a lot of opportunity to really shorten those time lines, and that's being done, right? So the in silico prediction is basically computer analysis, right? So in the right having -- with the right resources, this could be done in like a week or 2, we -- when -- in our -- certainly an academic efforts like no question. But even in some of the industry efforts like it takes much longer than it sort of that needs to be taken. There's a lot of like operational logistical slack that can be removed and that's when an opportunity comes in, I think so. I think in an ideal world, you should be able to predict the epitopes in like a week, say, and then manufacturing is a little bit -- that obviously depends on what type of vaccine you make. But even there, at least from what I'm -- for example, the most familiar with peptide vaccines, there's definitely opportunity to speed that process up as well. I mean it typically takes like several weeks now, but there's no reason this needs to take more than like a week or 2. That's like take that a little bit with a grain of salt because certainly not like sort of in -- or somebody who's like really involved in the nuances of making the peptide or even like let alone like an RNA vaccine. But I think the bigger point is like there is not a built-in like time lag such as growing up T cells in the lab, which is like, but cannot, right? That's not something you can speed up versus like whether the personal -- the vaccine there is like pretty much every step of the way, you can save time.

Okay. With that, in the interest of time, we covered most of the questions in the queue. But Erik, if there's anything you wanted to add to Dr. Ott's comments on where Evaxion is in terms of speed, I'll leave it to you. And Niels to wrap up this session.

Maybe just a quick comment, and then I'll let Niels conclude. I think we share your ambitions, Patrick, in terms of speeding up the process. I think we've been very successful so far in really cutting down the processing time certainly in terms of identifying the new epitopes, I'm right along with you and your predictions. And again, as I mentioned, we have, I believe, an industry-leading production time, which we are down to about 6, 7, 8 weeks depending on the platform peptides or DNA, I think that's pretty good. And that includes worldwide shipping. But of course, we will see further reduction in those processing time lines. And that's -- we all harbor that ambition. So with that, I think I'll leave it to you, Niels, and thank you, everyone.

Thanks, Erik. And first of all, thank you, Dr. Weber, and thank you, Dr. Ott, for educating us all a little bit in this important field where still a lot can be done. I think that's one of the key conclusions. Immunotherapy has done a lot for these cancer patients, certainly with melanoma, for the last few years, but still there is a high medical need and there's a long way to go. I'm pretty sure that we can all agree on that. And that's why we at Evaxion everyday work to make sure that the immunotherapies that we developed are based on the PIONEER platform for cancer neoepitope selection is optimized. That's an ongoing effort. We also work with different delivery modalities to make sure that the neoepitopes or the neoantigens are delivered to patients in the appropriate way because CD4 cells matter, CD8 cells matter. We don't have the clear answer yet, but we need to test our way and that's part of our ambition to test our way and make our decision on an as informed basis as possible. I think what we're seeing with EVX-01, a peptide-based approach, is encouraging. And as we generate new data points in patients, we can continue educate our platforms to make better predictions, better decisions for us in terms of the target selection. So every data point is important, and I think that's one of the true and really exciting aspects of using machine learning platforms in the development and discovery basically of new immunotherapies. And if this concept works, as we heard also from Dr. Ott, it can potentially be used in various cancers, not only melanoma, but also lung and bladder and many, many other types of cancers in the future. So with that, I want to thank everyone for joining, in particular, Dr. Weber and Dr. Ott for making us all a little bit wiser today.

Thank you, everyone.

Take care.