Evaxion Biotech A/S - ADR Aktie

I'm Birgitte Rono, CSO and COO at Evaxion. So on behalf of the entire Evaxion team, I would like to welcome you to this exciting webinar focusing on our EVX-01 program in advanced melanoma and the Phase II results presented on Saturday at the ESMO Annual Congress.

I'm delighted to have Professor Adnan Khattak with us today. He led the EVX-01 Phase II study at the clinical site in Perth, Australia, and he will be presenting the data from the EVX-02 study. So Professor Khattak is a medical oncologist at Hollywood Private Hospital in Perth and is a key investigator on many international cancer trials assessing novel cancer immunotherapies, including personalized cancer vaccines.

Before we start, just a short reminder, today's presentation may contain forward-looking statements, which are statements about the future. Actual results may differ materially from them. So here is what we'll cover today. First, I'll introduce Evaxion, the EVX-01 concept and the results from the EVX-01 Phase I study conducted prior to the Phase II study. So second, Professor Khattak will introduce us to the unmet medical need in advanced melanoma, and then he will present the data from the EVX-01 study. Then in the end, we will open the floor for your questions. You can type them into the Teams chat at any time, and we will answer as many of your questions as possible at the end of the session.

Yes. So first, an introduction to EVX-01 and the key data from the EVX-01 Phase I study. So Evaxion was founded in 2008 with one aim: to decode the immune system with AI to address serious unmet medical needs. We are based in Denmark, and our team of more than 40 experts cover the whole value chain from target discovery to clinical development. And today, Evaxion is a pioneering clinically validated tech-bio company. This means that we work in the intersection of life sciences and advanced data technology. AI machine learning and computation are deeply built into how we discover and develop novel therapies.

Our AI platform is called AI-Immunology, and it is an integrated set of models for rapid, high-quality target discovery, drug design and development. We are focusing on antigen discovery and product development in cancer and infectious diseases, and we recently expanded into autoimmune diseases. We have been listed on the Nasdaq Capital Markets since 2021 under the ticker EVAX.

So AI-Immunology allows for rapid development of product candidates with all of our pipeline candidates designed with AI-Immunology. So the focus of today's webinar is on our lead program, EVX-01, and the Phase II data we presented at ESMO. So EVX-01 is a personalized cancer vaccine, and we designed a unique vaccine for every single patient in our trial. It is based on up to 10 different targets, so new neoantigen that arise from mutations specific to the patient's tumor. And these neoantigen are selected by our AI-Immunology platform and the vaccine is a peptide liposomal formulation.

And let me show you what is behind our AI-Immunology platform. AI-Immunology has a unique building block architecture. At its core are 3 engines: the first is disease decoding, which identifies key pathogenic drivers in a given disease. The second is immune response decoding, which shows how the immune system sees and responds to potential targets. And the third is a vaccine design engine, which is -- which we use to design the product candidates. And we can apply the platform across disease areas. And currently, we work within cancer, infectious disease and autoimmune diseases. And EVX-01 is one of the cancer applications of this architecture. But AI on its own does not give you a therapy. And that is why we have built multidisciplinary R&D capabilities around the AI-Immunology platform. And these different capabilities include disease biology, target discovery and further also preclinical development, CMC and clinical development.

So now to the concept behind EVX-01. EVX-01 is a personalized cancer vaccine targeting multiple private neoantigens. Neoantigens are ideal cancer vaccine targets as they arrive from cancer-specific DNA mutations exclusively found in tumors and absent from normal tissue. They trigger strong, highly tumor-specific immune responses and they work well together with current cancer immunotherapies. And AI-Immunology models the neoantigen process inside the cancer cells. This starts with identifying these cancer-specific mutations, then it looks at the expression and translation of these sequences. It goes on to look at presentation on MHC molecule and the T-cell responses and further into the clonality of these neoantigens.

The AI-Immunology platform is part of a larger process that runs from mining the patient's tumor fingerprint to designing a personalized treatment. For every patient enrolled in our trial, we collect a tumor biopsy and normal blood sample and then we sequence the DNA and the RNA. And this information is fed into our AI-Immunology platform, which then identifies and ranks the neoantigen according to their predicted likelihood of inducing a tumor-specific T-cell response. The prediction only takes 24 hours and AI-Immunology currently predicts immunogenic neoantigens with an 89% precision, and we'll come back to this target precision a bit later in the presentation.

The vaccine is then manufactured and administered to the patients in combination with checkpoint inhibitors. The turnaround time from the time from biopsy to vaccine administration is about 7 weeks. We tested this concept first in a Phase I study, and it enrolled patients with unresectable stage III or metastatic stage IV melanoma. In the Phase I study, each patient received a vaccine of up to 10 patient-specific neoantigens, again, designed with AI-Immunology. The patients were dosed with EVX-01 to activate the T-cells together with a checkpoint inhibitor according to label.

The Phase I study tested 3 different EVX-01 dose level. And these are the key results from this Phase I study. So EVX-01 was safe and well tolerated with only grade 1 and 2 adverse events reported. Eight out of the 12 enrolled patients responded to treatment with an objective response rate of 67%. And EVX-01 induced a specific immune response in all these 12 patients. We also looked into the individual ability of the neoantigens to induce a T-cell response, and we saw that we had the activity against 53 of the 91 neoantigen analyzed across the patients with a target precision rate of 58% and manufacturing was efficient with a turnaround time of approximately 7 weeks. We took 3 learnings from this trial into Phase II. We selected the highest dose level of 200 microgram per peptide as the recommended Phase II dose. We added EVX-01 boosters to sustain the T-cell level, and we improved the prediction power by updating AI-Immunology. So all 3 shaped the Phase II study.

And with this introduction, I'm now happy to hand over to Professor Khattak to present the EVX-01 Phase II data. Please go ahead.

Thank you, Birgitte. Thank you, everyone, for joining online. Okay. So today, we're specifically talking about advanced melanoma where the role for Evaxion vaccine has been tried in, but the discussion, obviously, thinking ahead can be obviously rolled out to a much broader level across other tumor types as well based on some recent data.

So when we talk to our patients in our clinic, our current standard of care is immune checkpoint inhibitor treatment as a single agent or as doublet immunotherapy in the form of ipilimumab and nivolumab combination or nivolumab and relatlimab combination. And as you can see, the 5-year survival statistics -- now you've got 10-year survival statistics as well -- but roughly, you can expect about 50% of your patients to be alive at 5-year time point despite having a diagnosis of metastatic disease. But then the problem is what about the rest of the 50%? And if I want to see a patient in the clinic today, I can share this data with them, but I can't tell them whether they're going to be in the 50% of patients who are likely to live or they're going to be in those 50% that unfortunately do not benefit from checkpoint inhibitor treatment.

Now these treatments are very generic, very broad-brushed approach. So if I get, say, an advanced melanoma patients, my standard of care treatment options that I'll discuss with them will be pretty much the same in terms of these agents that I just listed. So with the unmet needs that obviously, we have got a significant proportion of patients that who would not benefit from our standard checkpoint inhibitor treatment as single agent or even double agent as well, and obviously, there's significant toxicity as well.

So the field in oncology moved obviously to personalized treatment, but that was primarily the personalized targeted therapy, what we saw in melanoma as BRAF-targeted therapy or EGFR, ALK or other target therapies in lung cancer. But obviously, running in the immune system at a personalized level has always been a challenge and remains a challenge, although we made significant advances, particularly in the last 2 to 3 years, including the data that was presented at ESMO recently.

So in terms of personalized treatments now, as Birgitte mentioned, how the neoantigens work, they're just like the fingerprint of cancer cell, which is unique to an individual patient's cancer that is different than any other patient on a particular clinical trial that we are enrolling at that particular time. So these are called neoantigens, what we call -- what we used to call personalized cancer vaccine, and now we call them INTs or individualized neoantigen therapies. And the data that was presented at ESMO from Evaxion, from the Presidential session, from Moderna data and some trials that have completed recruitment, hopefully, we'll see more and more success of these new novel therapies.

So I'll move to the next slide. So in the Phase II study last year, in Berlin, we presented the 2-year data and now in Madrid we presented the 3-year data. So the objective response rate, I'll discuss the design of the clinical trial in the next slide, but in summary, the objective response rate was 75%, means 3 out of 4 patients benefited. We enrolled 16 patients in this Phase II trial. And then what we also looked at is that after the run-in of 3 months of treatment with pembrolizumab, once the Evaxion vaccine was started, we looked also at the conversion rate of the best response to pembrolizumab to see if we can further improve on that by exploiting the synergy between the checkpoint inhibitor and the Evaxion vaccine, so EVX-01.

And what we found was that in 4 out of 6 patients, about 67%, the response converted from stable disease to partial response. And in about 43% of the cases, 3 out of 7 patients it converted from a partial response to a complete response. And that response was maintained at 2-year follow-up in 11 out of 12 patients. We also saw vaccine-induced specific T-cell responses in all patients receiving at least 6 vaccinations in about 15 patients. What we saw in the 3-year data saw 6 patients consented in the extension phase, so I'll cover that in a second. So probably move on to the next slide that will go through it nicely.

So the Phase II design, pretty simple and straightforward, very clean design. So it took patients who were treatment-naive in unresectable stage III or advanced metastatic melanoma who were checkpoint inhibitor-naive or had a good performance status and RECIST measurable disease with no active brain metastases. The primary endpoint was obviously the conversion rate. So a patient in terms of CR or PR from patients with stable disease and also conversion from partial response to a complete response, with a secondary endpoint of objective response rate, progression-free survival, oral survival, safety and tolerability data and also the immunogenicity data in terms of neoantigen-specific T-cell responses.

So the patients would generally -- would normally start the treatment at the standard of care, which was pembrolizumab at that stage, now obviously, will be using doublet therapies, but at that stage it was pembrolizumab when the trial protocol was written. And then after 12 weeks run-in of pembrolizumab monotherapy, patients had scanned, and we noted that and then had 6 priming doses. So 2 weeks apart from week 12 to week 22, this is a priming phase and then there was a boosting phase at week 30, 42, 54, and 78. And then the trial was followed up till week 102 and then 90 days follow-up. And more recently, there was a protocol amendment in which 2 extension phase doses were given at week 115 and week 139 and then the end of the trial.

So 21 patients were screened, 4 were screen failures, 17 were enrolled, one progressed unfortunately prior to starting the Evaxion vaccine, and then 16 patients were recruited in the trial and then 6 patients entered into the extension phase of the study. Roughly 2/3 of the patients were male, almost a 50-50 split between less than 65 and greater than 65, so both young -- relatively younger and older patients were treated and patients with good performance status, which you'd expect in the clinical trial. And then the median number of target lesion was 2. PD-L1 expression less than 5%, greater than 5% in 25% and 50% of patients, respectively. About 50% of the patients have BRAF disease, positive disease and LDH was elevated in about 13% of patients.

In terms of safety first, because Phase I study first need to demonstrate the safety of the study -- sorry, Phase I leading into Phase II later on. So we only saw mild grade 1 to 2 toxicity. So if you look at -- so these are the grade 1 toxicity, grade 1, 2, 3, 4, 5 and then attributed to Evaxion only, you can see there's nothing in grade 3 and 4, only grade 1 and 2 toxicity seen. And same with pembro as well, what we expect with a normal profile, except one grade 4 immune-related diabetes that was seen from pembrolizumab effect. And then no safety signals were seen for the combination as such.

In terms of the efficacy of the treatment, the objective response rate was 75% and 25% complete response rate. You can see the waterfall plot. So this is a patient with progressive disease. Apart from this 1 patient, everybody else had at least some degree of reduction in the tumor volume, including 4 patients with complete response. The blue bars representing partial response and the gray bar representing patients with stable disease. And then 92% of the patients who responded demonstrated sustained response at 2 years of follow-up.

Going into the tumor kinetics. So you see that the green lines in the spider plot representing patients with a complete response, blue line representing patients with partial response, gray line patients with stable disease and then 1 patient who had increase in the tumor burden, as you can see in the red lines. So in terms of complete response we discussed 25%, partial response 50%, giving us a total of 75%. But what I find more interesting from a clinician perspective is this diagram. So these are all the patients. So the red circle is a time point when they had a response assessment. This is the first response assessment, which was done at week 12 after starting immunotherapy with pembrolizumab. So these patients did not have any vaccine at this stage. And then you can see the priming doses with these little arrows concentrated together, and then we got the 4 booster doses.

The green represent a patient with a complete response, blue partial response, gray stable disease and red with progressing disease. What is interesting is if you look at, for example, this patient who had a complete response, you would expect that, okay, yes, they had a complete response to pembrolizumab that was continued. So you can't argue that the vaccine necessarily improved anything further because they already had a complete response.

But then if you look at patient number -- this patient and then this patient and this patient, you can see these patients already had pembrolizumab for 3 months. And then further -- and then later on, you can see when the effect of the vaccine really kicks in, you can see that these patients start to convert from partial responders to complete responders, the blue turning into green. And then some of the patients with stable disease, these patient had stable disease, and then converted into partial responders. And particularly when you look at these patients, stable disease, stable disease and then converted into partial response.

One of the questions that was asked, I think, at ESMO last year was as well is that these patients were already responding to pembrolizumab, so maybe that was just consolidated. But generally, we expect that most of the responses are relatively, clear in the first 3 to 4 months after starting pembrolizumab. And even when you look at these patients, these are quite far down the track at almost week 46 into treatment. So normally, in clinical practice, we would have seen that they had a response -- the best response would have been seen to pembrolizumab by that time.

So in terms of overall improvement, as I mentioned in the summary slide earlier, so overall improvement in response from stable disease or partial response at week 12. So 7 out of 13 patients had improvement in their response rate on top of pembrolizumab, stable disease to partial response in 67% of patients, partial response to CR in about 43% of patients. Objective response, obviously, 75%. Median overall survival and progression-free survival has not been reached.

And then the clinical responses of the 6 patient extension phase, they also continued to respond and they also continued to demonstrate the activity. So you can see that Evaxion's AI platform identified with 89% success rate of the immunogenic vaccine neoantigens and then 86% of them being de novo neoantigen response that was not driven by pembrolizumab. And if you look at this graph, so we got the EVX-01-specific T-cell responses measured by interferon gamma ELISpot assay. And you can see this is the prevaccine when only pembrolizumab was given, then the priming phase, and then the boosting phase, and then the extension phase.

And you can see, as soon as the pembrolizumab is given, so we're not really seeing a spike in the T-cell responses. And then when the vaccine is given, you can see the amplified neoantigen-specific vaccine-specific T-cell responses and the effect of the boosting dose, and that's sustained in the extension phase, which is very encouraging. So when you look at the data in further details, so we saw both CD4 responses and CD8 T-cell responses, which is again very encouraging. Now some of these patients, the data is not available yet. So we'll find out when all the samples are analyzed, and we might see more CD8 T-cell responses in these patient that we haven't ticked yet. And you can see this patient #12, you can see clearly CD4 and CD8 T-cell responses before vaccine. And then at week 30, you can see clearly the responses seen at immunological level.

So in conclusion, so we saw a good objective response rate of 75%. Yes, the sample size is small. We saw some preliminary evidence that we could potentially consolidate further and improve further on the responses achieved by checkpoint inhibitor treatment and vaccine-induced T-cell responses were sustained and then there was a durable benefit seen clinically as well as looking at the immunogenicity data, particularly from the extension phase, none of these patients who went into the extension phase has had disease progression and the vaccine [indiscernible] success rate of 89% is very encouraging. Thank you.

Thank you so much, Professor Adnan, for sharing your perspectives on our data. So we will now open the session for questions and please type them into the Teams chat and I'll just give you 1 minute to start asking or typing the questions in the chat.

Birgitte, can you see the chat box because I can't see it.

I can see the chat box in the upper ribbon, there is a chat function. If you click on that, you should be able to see the questions. I can read them if that works for you.

I'm in the chat box. So I think maybe if you can read the question and then I'll try to answer it.

If you scroll down, you see the questions. Do you see them?

Sorry, I think if you maybe start -- bring up the question, and then I can try to answer. Yes.

Yes. So first question is from [ Miguel Longobardi ]. So hello, hello Birgitte and Professor Khattak. [ Miguel Longobardi ] here. It was a pleasure meeting you both recently. I have 2 questions for today's +discussion. So for Professor Khattak. What evidence most strongly support EVX-01's contribution to durable response or disease control after pembrolizumab discontinuation?

So I think in clinical practice, obviously, it's harder to tease out exactly what drug is contributed because obviously, we do see with checkpoint inhibitor that you don't need to continue the checkpoint inhibitor treatment beyond a certain time point. The trial protocol stipulated 2 years of treatment. But obviously, in clinical practice, we also see patients who continue to drive benefit even if we stop their treatment at 6 months, sometimes 12 months. So the checkpoint inhibitor response can continue.

Additionally, if you look at the immunogenicity data and the T-cell response data that I just showed on the slide, so we can still see that there's vaccine-specific T-cell responses going and there's also a corresponding clinical responses ongoing [indiscernible] these patients, particularly in the extension phase have serious disease recurrence. And that I think further obviously, also comes from the conversion data that, yes, I think pembrolizumab did the heavy lifting to begin with. But I think after 3, 4 months, I think generally, we would expect most of the responses to plateau rather than continue to improve.

Yes, we see some late responders coming through, but beyond 5, 6 months, it is a bit less common. So most of the routine clinical practice scans that we look at, usually in the first or the second scan, you'll be able to see most of the responses. And then I think when you look at those dotted lines, some of these scans were happening at week 48. And at that time, they were converting into their best response from stable -- sorry, from partial response to a complete response or down the track from stable disease to a partial response as well.

So yes, I think, at molecular level, we can't say that this is definitely not related to pembrolizumab. But I think clearly, we see that when you look at it that we -- from a clinical practice, we know that we're not expecting too many late responses. On top of that, we see that there's a spike in the antigen T-cell response, CD4 and CD8 as well, looking at the immunogenicity data, and that's because when the effect of the vaccine really kicks in. So from what we know, for example, from the Moderna KEYNOTE-942 and the INTerpath trial data, the curves start to separate from 6 months and beyond, and that's when the priming doses have happened in effect and the T-cell responses really kick in. So I suspect that's what we're seeing as well.

And I think from clinician perspective, I think scientifically, it makes more sense that probably most of the late responses of conversion is happening from -- or to a certain degree, definitely from the Evaxion vaccine as well.

And then there is a second part of the question from Miguel. So given EVX-01's 89% neoantigen immunogenicity rate and Moderna-Merck's positive Phase III results, which is EVX-01 strongest potential competitive advantage and what further evidence would a pharmaceutical partner need?

So I can give that a go. So there are, of course, several differences between EVX-01 and Moderna-Merck's therapy. So first of all, EVX-01 is a peptide-based vaccine. And we know that peptides are immediately active upon administration, whereas the mRNA molecules, they need to be translated. There's also differences in terms of the stability of these different vaccines. So peptides are normally quite stable, whereas RNA molecules are prone to be degraded by different enzymes in our environment, and they also require cold chain supply, which can be quite, I would say, challenging.

We have a very high immunogenicity rate. So 89% of our neoantigen targets across the patients in the Phase II study are giving rise to a strong and specific T-cell response. We have not seen a lot of data from Moderna-Merck in terms of T-cell responses. At ASCO this year, they reported data from 7 patients in their Phase II trial and the immunogenicity rate or nonantigen hit rate, as we also call it, is significantly lower than what we have reported in this Phase II trial. So approximately 30% of their targets are giving rise to a T-cell response. And we believe that this is a clear testament of the precision and predictive power of our AI-Immunology platform.

Further, we cannot compare the clinical data directly between the 2 programs. EVX-01 Phase II study was in advanced melanoma, whereas the Moderna-Merck trial is in a different patient population, so in resectable melanoma. So there are definite differences also in terms of safety profiles. As Dr. Khattak just alluded to, we only have grade 1 and grade 2 adverse events, whereas there was a little bit higher or more severe adverse events reported from Moderna-Merck, not anything that could not be managed, but they do see some grade 3 adverse events.

Thank you, Birgitte. So I'll also probably like to build up on that. So these are obviously 2 different ways of administering personalized neoantigen therapy, which Birgitte very nicely summarized, the peptide vaccine versus mRNA vaccine. And obviously, the patient context is different. So the Moderna data is purely from high-risk resected melanoma patients in the Phase II study and also in the Phase III study as well. So they are primarily stage II, III and some resected stage IV patients. So the tumor microenvironment is expected to be different in that context compared to late-stage disease, which is Evaxion study. And I think Birgitte is absolutely spot on as well that there's limited translational data from the KEYNOTE-942 Phase II study. We probably might get some better quality data from the INTerpath-01 clinical trial down the track, but it will take some time because we've just seen the preliminary results.

In terms of -- I have been involved with both those trials and have been involved with the Evaxion trials as well. In terms of patient feedback, I think the personalized cancer vaccine from Moderna, I think definitely, I thought it was a bit difficult to tolerate in terms of the vaccine-specific adverse events and the flu-like symptoms was more with mRNA vaccine and much easier to give to the patients Evaxion vaccine. And I think Merck is also trying, obviously, the INTerpath-12 studies currently running. So pembrolizumab versus pembrolizumab plus the mRNA vaccine. It's recruiting slowly. It's got a difficult inclusion/exclusion criteria, but that will probably provide us some more evidence as well in terms of how the Evaxion data will compare with the mRNA data in the metastatic setting.

So the next question is -- I can see the question now from Thomas Flaten. So thank you -- yes. Yes, Birgitte?

Please go ahead.

Okay. So if this were Phase III data and formed the basis for regulatory approval, where in lines of treatment will you see Evaxion.

So I think as a clinician, I can see Evaxion's location in multiple -- obviously, the data is supporting advanced disease. So we don't have 3-year sustained personalized cancer vaccine data from any other vaccine at this stage in the metastatic setting, so thus Evaxion is winner by miles in that. And in addition to that, we got sustained immunogenicity data to support its use as well and with a high precision rate of neoantigen selection. So Moderna one was 34, we got 10 here, and the data that we have presented here is also [indiscernible] that we can refine it further by a PIONEER score, which is like a biomarker where you really pick up the top 10 neoantigens. So you enrich your population, enrich your neoantigens who are more likely to respond.

Where will we use it? So I think it plays both in the early and late stage. In the late stage, obviously, we don't have data for ipi/nivo or nivo/rela in combination with Evaxion because that is the preferred standard of care for most patients, but we still see a lot of single-agent use in routine practice because not everybody is fit enough to face the toxicity of the ipi/nivo combination. So that's why I think in Merck's study as well, pembro is the backbone.

So definitely, I think with a single agent in the metastatic setting, starting off with a checkpoint inhibitor treatment and then adding in Evaxion is the way to go, although we probably won't put patients with high disease burden that are rapidly growing because they might progress faster and not well enough to be able to tolerate by the time the vaccine is manufactured 6, 8 weeks later. The disease might be raising away quickly. So generally, we'll put patients with a relatively low disease burden.

And then also, you can start up with a checkpoint inhibitor and those who do not respond, then you can use it to rescue them potentially, which we did in this trial, not necessarily as a rescue, but trying to further consolidate what we can achieve with a single agent. I don't see any reason why it can't be looked at in the combination with doublet as well. And then in terms of early stage, obviously, in routine practice now, what I found is that we're seeing less and less advanced melanoma patients because we're not using the checkpoint inhibitor in the stage I and stage II setting. We're using a lot of neoadjuvant treatments as well, which is complicating how we park these different treatments.

One other exciting area where we could park this is potentially in the neoadjuvant setting, but more in terms of those who do not respond to the neoadjuvant treatment, whether we can rescue that group of patients because if you look at the NADINA neoadjuvant trial data, you look at major pathological response rate of about 60%. This means about just under half of your patients are not driving good clinical benefit. So using it for those patients in the adjuvant setting rather than giving it to everybody. So that will be one field as well. And then looking at other types like mucosal melanoma or [ rectal ] melanoma, which are less susceptible to treatment with checkpoint inhibitor treatment. So that will be another area to look at potentially.

So I think, we skipped the first question in the chat, not on purpose, but maybe we should just go back to that question from Michael Okunewitch. Can you see that, Adnan?

Yes. So I think 10 neoantigen is fairly low compared to personalized vaccine. Why not target more and what in particular gives you the confidence? I will start at this, but I think it doesn't work that way that if we have 100 neoantigens, so that will lead to success or 10 is inferior compared to 34. So I think what is important is that 89% success rate in terms of generating a vaccine-specific T-cell response, that is more important. You can load 34, and that's the maximum that can be done in the Moderna's case. This doesn't mean all 34 of them are immunogenic. And we do not have the data and integrated data in terms of how much of a proportion were immunogenic.

So I think the 10 neoantigen is which are specifically selected by the AI platform that are most likely to be immunogenic, which is supported by the data for 89%. And Birgitte, you might have further to add to it as well at the molecular level.

Yes. So we are definitely in favor of quality over quantity. And the reason for that is that we don't want to administer neoantigens that are not driving a strong immune response. Whether 10 is the ideal number, we don't know. 15 could also be the case. What we do know is that we would like to target the tumor with multiple different neoantigens. There is evidence in the literature of tumor escape, et cetera. So if you hit only 1 target, there might be clones that are outgrowing that does not contain this single neoantigen.

So mult-targeting is important. It's also important that we choose the right neoantigens and that we don't dilute or disturb the immune system. So that's the approach we have been taking. And I think the 89% target precision speaks for itself. It's definitely a testament to the precision of our AI-Immunology platform. We don't know how many is needed in terms of driving a strong clinical response. But we believe that a strong and specific immune response is the key factor here.

There's also a second part of this question, Adnan, and I think that is more down your alley. So what degree of relapse would you normally expect in the year after checkpoints stop?

Yes. I think good question. It depends whether the treatment was stopped due to toxicity and what degree of best response we got from the checkpoint inhibitor because in clinical practice, we would normally stop treatment sometimes if the patient had toxicity or if they had a complete response or sometimes even if they had a partial response as well.

But generally, I think if you look at the data, I think there's some fresh data on that as well presented by Caroline Robert that if patients had voluntarily stopped treatment due to either toxicity or due to complete response, and when you look at particularly patients where they had a good response and the treatment stopped, you would expect probably somewhere around about 15% to 20% potentially recur. And then not all of them will respond to treatment on rechallenge. So I think it's a bit complex to answer this because it depends upon what happened and what was the response, but it is somewhere around about 15% to 20% number.

And then I think we answered Miguel's question, and we answered Thomas' question. And then Patrick, for the future of EVX-01 is a major partner required? I think the answer to that is, yes, I think ideally, we would like to have a major partner in terms of the immune checkpoint inhibitor treatment. Obviously, we collaborated with Merck on this study. But certainly, I think a major partner is required and obviously to further build on the data and explore other avenues of future use of Evaxion, not only in melanoma, but potentially in other tumor types as well.

Yes. So the strategy of the company is definitely to out-license our product candidates, including EVX-01. And there is, of course, a need for further clinical development also on EVX-01. As Adnan is alluding to, we are also looking at other indications where we definitely see a fit for our therapy, and that includes different indications where there is a high mutational burden and where checkpoint inhibitors is approved, but there is still a need for adding additional therapies. So we see a fit for other indications beyond advanced melanoma.

Okay. So next question is, what is your perspective on developing peptide DNA or RNA-based vaccines? Would you consider exploring RNA-based vaccines for different indications? It's a good question. I think absolutely, the field will really take off, I think, based on the research that's going into now trying to personalize the immune modulating treatments rather than using more generic off-the-shelf immune modulating treatments, although they're probably likely to be used in conjunction, I think, for quite some time.

We've seen the RNA vaccine data from Moderna, the Phase II and now the Phase III data, and then we got the Evaxion data as well and some early biotech data as well. So certainly, the field is likely to head this way. And then rather than one size fit everyone, we will have more in terms of one drug for each patient in combination with checkpoint inhibitor treatment and certainly exploring in other tumor types as well. We haven't looked at Evaxion's use in other types yet, but it is being discussed internally between the Evaxion team.

But in terms of the mRNA-based vaccine, certainly, I have participated in trials of INTerpath-1; results were just presented. INTerpath-2 has nearly completed recruitment for adjuvant lung cancer, INTerpath-5, which is adjuvant bladder cancer with checkpoint inhibitor with Moderna's mRNA vaccine. INTerpath-7, which was in squamous cell carcinoma got rolled back. And then we are actively participating in a stage 1 high-risk lung cancer as a preventative strategy for high-risk stage I lung cancer, pembrolizumab with the mRNA vaccine as well.

So across lung cancer, bladder cancer, and then INTerpath-4 RCC adjuvant study that has completed recruitment. So some Phase III trials have already completed recruitment and probably get the data in the next couple of years and some are actively recruiting as we speak. So the field is really exploding, and we're all excited as oncologists and clinicians and collaborating with our pharmaceutical industry colleagues as well as translational research colleagues as well.

And maybe I can add to that. So we are, of course, all of us very excited about all the new data that will come out in the coming years. In terms of the modalities, so peptides versus DNA RNA-based vaccines, I think we can conclude that the field does not know currently which is the preferred modality -- we have been exploring peptides and DNA-based vaccines, both in the personalized space and now advancing into off-the-shelf therapies based on a DNA vector. But there are no direct comparisons between these different modalities. We have done a little bit preclinically, but whether that would translate into clinical settings, we, of course, do not know.

From an immunological point of view, peptide-based vaccines are most likely prone to drive a strong CD4 response, whereas DNA and RNA have a more skew towards CD8 T-cell responses. And we believe that we definitely need both CD4s and CD8 T-cell responses in order to drive a strong response in the patients. But the field has not settled completely on whether DNA, RNA or peptide-based formulations are the most favorable ones.

Thank you. So next question is, how does the evolving standard of care in melanoma change the opportunity for Evaxion? Does the availability of combination to induce stronger responses make Evaxion more attractive to -- for durability of response now that the 3-year data is durable?

So doing trials in melanoma is very attractive, but at the same time, very challenging as well because the field is moving so fast. So a typical trial protocol usually is written up to maybe at least 1 to 2 years before a trial is really rolled out, the final protocol is agreed on, and then by the time you take it through the regulatory authorities, start-ups happen, recruitment happens and the standard of care changes. So it is always very difficult to obviously stay up-to-speed with the clinical trial landscape that is changing rapidly.

Yes, certainly, I think our data is just pembro with Evaxion, but I don't see any reason why the same can't be done in combination with combination treatments like nivo/rela or ipilimumab/nivolumab and particularly, I think, because not everybody would respond to these drugs either. So I think the first slide that I showed you, 50% survival rate was with combination ipi/nivo. So this means 50% do not benefit. So certainly, it can be used in combination with those patients as well because -- and also on top of the checkpoint inhibitors, I do not believe that Evaxion -- I can't assume for sure because it hasn't been done in the trial, but I wouldn't expect that the peptide vaccine would add significantly in terms of the toxicity on top of what ipi/nivo or nivo/rela might do. So certainly, it can be explored in that context.

Next question is, could this be used in newly diagnosed melanoma patients with lymph node metastasis. So a stage III disease high risk, like the INTerpath-01 trial with mRNA vaccine was positive. Certainly, this can be looked at in the early-stage setting as well. But then you can argue that, obviously, one vaccine has already demonstrated improvement, doing the same thing from a clinical perspective, yes, can be done. And obviously, the differences between peptide vaccines in terms of the logistics, the stability, the cold chain stability, some advantages of one over the other and the timing.

And then also, as I mentioned in the earlier question that in the neoadjuvant space, which is also being increasingly popular almost for 50% of the patients that we see with stage III disease with lymph node metastases, they've got microscopic disease. And then obviously, 40% -- somewhere between 40% to 50% of them do not respond or do not get a major response, so it can be used in that context as well. So some neoadjuvant trials are being planned and rolled out. It will be nice if somebody can look into that aspect as well.

If you were designing a Phase III clinical study, what would be your recommended control arm? Pembro plus ipi, for example. Obviously I need to think about it with a more clear head. But yes, I think one of the major challenge is the question now asked, for example, I did this Evaxion study, same question is asked, why not ipi/nivo. But obviously, by the time we were doing it, ipi/nivo is still not available in many countries in the world. And even in Europe, I think it just -- I think at least German colleagues tell me they just got access to it in the last, I think, 12 months or so.

So I think data is one thing and availability is a different thing. And then ipi/nivo generally would be considered or nivo/rela would be considered now the active comparator in the contemporary trials. But we've got many trials that I've participated. I still am involved. For example, we're doing a TIL therapy trial. That is also pembro versus pembro plus TIL. And one of the major criticism is that why pembro. But obviously, when the trial was designed, that was the standard of care.

And we also have to bear in mind, obviously, that there has been no clear superiority of ipi/nivo versus nivo, for example, checkpoint inhibitor treatment in the CheckMate-069 trial because the trial was not statistically designed to prove superiority or didn't have a statistical power to prove the superiority of doublet versus single agent. We think that maybe doublet has got an edge, but it wasn't statistically proven so.

And then the last question on behalf of R.K. from H.C. Wainwright. So I think Evaxion's data is obviously a small study in the metastatic setting. So it's a different context. The data from Merck-Moderna is in a different context in early stage with a different tumor microenvironment. The pros and cons of the vaccine delivery between peptides and mRNA, which I think Birgitte answered already. So I think eventually, we'll go with whatever the data supports.

Obviously, if there was an Evaxion trial in the metastatic setting, then obviously, if that was demonstrated pembro plus -- or whatever is the competitor arm plus Evaxion is better than the standard of care, then obviously, the clinician will follow that. So we go by data rather than necessarily specifically looking at individual nitty-gritty of the differences in the logistics. But I think logistic of maybe the peptide vaccine might have some advantages potentially and the patient tolerability was a bit better anecdotally.

But obviously, I haven't treated many patients with Evaxion because the overall number of patients is small, whereas I've treated hundreds of patients or not hundreds, but 100-plus patients with the mRNA vaccine. And certainly, it can be quite challenging to give to some of the patients because they get quite a marked flu-like illness. And clearly, some of the younger patients in the 40s and 50 who we expected to have stronger immune response de novo, they have a much stronger reaction to the extent that I had to give even low-dose prednisolone or even opioids to deal with the myalgia for some patients.

So I think the more we use these vaccines, we'll get more knowledge about what they can do and how best to mitigate with the toxicity as well. And then within what you can discuss publicly, the approaches differ meaningfully in tissue requirements, ability to manufacture a vaccine, injection-related symptoms.

So I think we discussed that in terms of the injection site reaction. But definitely, I think anecdotally, I felt that it was easier to treat Evaxion patients, but have treated many more, obviously, with mRNA vaccines. So probably the sample size is different. So we can't draw any meaningful conclusions regarding that.

In terms of the tissue requirements, there's a difference in the sense that we had much easier tissue available for the INTerpath trial because it was done in the postoperative phase where you have an axillary dissection done and a big sample provided whereas the Evaxion was fresh biopsies in a metastatic setting. So it was a pure biopsy rather than surgical sample. So I think the inherent design of the Merck trial was such that obviously, they got an advantage because they got a better and bigger tissue because tissue is the issue.

But we didn't have any vaccine failures, at least from our site. So, y Birgitte, you might know whether you had any vaccine failures overall or not?

No, we did not.

Yes. Which is again very encouraging because if you look at the Moderna data as well, so they had 99% success rate. So the vaccine failures are low. We had 2 or 3 patients from our site in the INTerpath trial with vaccine failures, not because the tissue wasn't enough. It's just that I think they just could not, I think, extract DNA to be able to make the vaccine.

And then from a translational perspective, what differences would you expect between long peptide plus liposomal adjuvant? Birgitte, do you want to take this one on #3, so the translational one?

Yes, definitely. So I think we covered some of this already. As I mentioned before, so the peptide-based vaccines, they are more prone to drive a strong CD4 response compared to the RNA and DNA-based vaccines. And that's simply due to the way that the immune system takes up these different molecules.

As I mentioned, we believe that there is a need for having both CD4s and CD8 T-cells. It's long been considered as CD8s are the one that drives the cytotoxicity towards actually killing the tumor cells, whereas the CD4s has historically been considered as helpers or supporters of this process. But recent evidence definitely demonstrates that CD4 T-cells to also have the capacity or the ability to drive a direct tumor killing effect. So I think there is -- again, the field doesn't exactly know which balance between the CD4s and the CD8s are the most favorable one. But from a pure immunological point of view, I think we can say that at least having both types of T-cells can drive a strong antitumor effect.

And then I think there was a question around have you seen evidence of a clinically relevant difference rather than simply different laboratory readouts. I think that's difficult because we have not done comparisons.

Yes. And also I think the CD4/8 responses, again, we haven't done comparison. And also, I think we don't have that enough granular data from the mRNA vaccine, Moderna vaccine. And obviously, the vaccine study is in the metastatic setting, so you can do some correlation with response, whereas the INTerpath data, the KEYNOTE-942 data is in the early-stage setting. So you can't determine the response because obviously, that's in the postoperative setting and you only find out if somebody relapses. You can't look at the success of the treatment, you only look at the failure of the treatment. So it's a bit difficult to tease out in that regard.

But I think in the coming years, we'll probably drill down deeper into the CD4, CD8 responses and see if there's any correlation with or preferential correlation with the response of one or the other. And I think I also mentioned that, yes, there's CD4 responses that we're seeing in every single patient, but there was a significant number of CD8 T-cell responses that we're also seeing in these patients and a few that were missing, there's still some samples that need to be analyzed yet.

And I think there was one part of RK's question two that we didn't entirely cover. So he's also asking about injection-related symptoms or missed doses or patient willingness to continue. Do you have any comments to this?

Yes. So in Evaxion trials, everybody got the vaccine and they tolerated very well and nobody questioned that I can't continue. In the INTerpath and the KEYNOTE-942 trial, some patients did have quite a strong flu-like illness and myalgias and fever-like syndrome. So that usually happens -- starts from up to 6 hours after the vaccine to 36 to 48 hours sometime. And if it happens every 3 weeks with the vaccine 9 doses that were given, so it was challenging for some patients.

But patients that are motivated as well. They want to -- obviously, they see that maybe we're seeing a stronger immune modulation going on, and that's why we're getting these symptoms. But there were some elderly patients who were quite concerned about continuing it because it was affecting them. And from speaking to other colleagues from other centers as well, some of them did withdraw.

Thank you for that. So I think we have managed to cover all the questions. And I would just like to conclude the session by thanking you, Adnan. It's been a pleasure. And I know you just arrived home in Perth in due time. So thank you very much for presenting our encouraging data and also giving your perspectives. And then I would also like to thank all the participants for joining this webinar. So thank you all.

Thank you.

Good day, and thank you for standing by. Welcome to the Evaxion Business Update and Second Quarter 2026 Financial Results. [Operator Instructions] Please be advised that today's conference is being recorded. I would now like to hand the conference over to your speaker today, Helen Tayton-Martin, CEO. Please go ahead.

Thank you, speaker. I'm Helen Tayton-Martin, I'm the Chief Executive of Evaxion, and we're delighted today to be building our Q2 business update. I'm joined today on the call by Birgitte Rono, our CFO and COO, who will provide an overview of our updates on our R&D pipeline and AI Immunology platform. Then I'll hand over to Thomas Schmidt, who will talk through our Q2 financial results. before we bring it back to conclusions and Q&A.

So first, of course, we may make forward-looking statements, and the audience is advised to look at our recently filed SEC documents. And now I'll kick off the discussion. So really, Q2 has been marked by a series of achievements in our 4 core areas of 4 core platforms for the company. First of all, just focusing on business development. As previously and ongoing through the course of this year, there are any discussions we are having with partners regarding the action programs and pipeline.

We've had a stream of very encouraging new data, which continues to come through and continuously validate the AI immunology panel, which really feeds into those various conversations, and we'll touch on some of those today. And in particular, in our R&D area, we have been very pleased to be accepted to present further updates on our Evaxion program. our personalized neoantigen cancer vaccine in advanced melanoma patients. And obviously, it was a great day for the field yesterday to see the Phase III -- positive Phase III results from the similar Moderna Merck program in personalized cancer vaccine in melanoma produced. And so it will be great for us and the field to talk more about that how we had to ESMO and an update on our own data there.

Elsewhere, we've been working to refocus and expand the pipeline, leveraging our learnings with our anti-1 platform actually into a new program, which we call EVX-05 in glioblastoma, where we are further leveraging the ores that we have been able to identify highly conserved antigens for glioblastoma, building on what we have done in our EVX-04 program using a similar approach to use their immunology to find highly conserved and [indiscernible] in AML. So we presented new preclinical data met earlier this year at the European Hematology Association Conference, Annual Conference. And we also updated in our infectious disease portfolio on EVX V1 CMV program to the recent HSV, [indiscernible] conference last month.

More broadly on AI Immunology, the platform itself, we were really delighted to see that recognized in the Gallian, a second Gallian award we have had for the technology in the last 12 months. So very exciting to see that being recognized more broadly more globally in terms of the value in AI immunology prediction for our programs in infectious disease, oncology and autoimmune disease. And finally, in terms of the core of our updates, we have maintained a disciplined focus on our resource allocation, really strongly aligned to where we can build the most weight from the platform.

And with that, plan we can confirm that our cash runway remains unchanged with the cash at hand to fund our operations into the second half of 2027, and Thomas will talk more about that. So just a reminder, before we jump into it, our pipeline consists of a number of programs in cancer and infectious disease at the current time. Evaxion will be a focus for [indiscernible] presentation in a few moments and obviously, also including our EVX-04 and EVX-05 programs, which are focused on the concerns of anti-off-the-shelf antigen vaccines.

For infectious diseases. We have a number of preclinical programs there and some of which are partnered 1 with Merck with Aprogen and data is continuing to build on the interest that we have on those programs from partners. So in terms of where we are as we meet the halfway point of 2026. We have already met the first of our milestones in terms of updating on the EVX-01 platform at AACR earlier this year with biomarker immunogenicity preclinical data alongside the clinical data from year 2 at ESMO last year.

We will -- we have mentioned already, and we will be updating on the 3-year data from that program with efficacy results in ESMO in October. And in the rest of this course of this year, we will be talking more about the application of AI immunology in autoimmune disease as well as planning for the regulatory filing of that EVX-04 program, the outer-shelf program in AML. And finally, we will have an update on our group [indiscernible] program in -- with the design and preclinical validation of antigens in the EVX-B4. And we continue to prosecute a partnership approach around these programs and platforms where we see value creation.

So with that, I'll hand over to Birgitte, who will talk you through our R&D and AI immunology update.

Thank you, Helen. So today, our focus on our lead asset, so that our personalized neoantigen cancer vaccine currently in place 2 in advanced melanoma. Then I'll present our new official 5 vaccine program. demonstrating the scalability of our AI immunology platform into the hard-to-treat and define brain cancer glioblastoma. So lastly, I'll showcase how the immunology identified T cell epitopes are relevant in controlling CMV infections. So as Helen mentioned, we will present 3-year effect 1 Phase II outcome data at the ASCO Congress in October, and this data includes evaluation of the vaccine's effect as stand-alone and also in combination with anti-PD-1 treatment.

The data will potentially give further insight into enhanced effect -- treatment effects and also the durability of EBX-01-induced immune responses. And collectively, these data provide a more comprehensive assessment of the potential of Evaxion, so strengthening the already strong clinical data page.

So looking back at previously announced data from the Evaxion Phase II trial, we reported strong EBX-01-induced immune activation at the AACR meeting in April. So we were able to show that 86% of the EVX-01 vaccine target triggered a tumor-specific immune response which is a substantially higher frequency than what has been reported for other similar vaccine candidates. Furthermore, we also showed that 86% of the immunogenic vaccine targets induced a Novotel response, meaning that EX1 specific triggers, novelties and responses rather than amplifying existing responses.

This is very important as induction of the novel T cell responses has been linked to clinical benefit. And at the ESMO Congress last year, we recorded 2-year outcome data, including a 75% overall response rate complete responses and 92 of the patients still being in response, indicating doable clinical benefit. And importantly, more than half of the patients converted into an improved clinical response or current EVX-01 treatment. So over the last approximately 10 years, personalized new stream vaccines has shown promise across several early based clinical studies.

And with the Moderna America announcement yesterday, we can definitely say that the field is moving from promising experimental immunotherapy towards a clinically validated therapeutic modality with a clear and realistic path to regulatory approval. And these data are not just only a wind from Moderna and Merck, but it's a win for the entire field as they broadly validate the personalized new antigen vaccine concept. So overall, with our encouraging EVX-01 data and with the validation from Modena and Merck, we believe that we are well positioned as we move forward towards credit creation.

And so let's turn our focus to our after-shelf cancer vaccine programs. So in collaboration with Duke University, we are developing an after shelf vaccine, EVX-05 for glioblastoma or GBM, targeting conserved antigens as announced earlier this week. So GBM is the most common and most aggressive primary malignant tumor brain and despite surgery followed by chemo radiation outcomes remain very true up with a median overall survival of approximately 1 year, underscoring a significant unmet medical need. So our EVX-05 approach builds on the same novel and broadly applicable concept as EVX-04 as it is assigned with AI immunology to target conserved tumor-specific antigens debarred from endogenous retrovirus lens or beers, which are part of the dark genome.

The target selection process allows for a broad tumor coverage despite immune and tumor erbantigen differences across patients. So we have applied AI immunology. So our AI-powered target discovery across and identified an optimal set of bar fragments based on trust patients relevance and immunogenic percentage. And we have mined patient sequencing data identifying approximately 1.5 million [indiscernible] and selected 16 of this as the fragments that will be included in the EVX-05 vaccine.

So next steps include lead candidate selection and IND enabling activities prior to a first in-human study that is expected to be conducted in collaboration with the world-leading GM experts, we are collaborating with the GC University. So our other after-shelf cancer vaccine program, EVX-04 is also progressing well. So EVX-01 targets in multiple concerts in the case of this program identified in AML patient samples.

So as Sean mentioned, we presented novel data at the European Hematology Associates Conference in June demonstrating that the EVX-04 vaccine is expressed and secreted by human cells, enabling immune recognition and activation. So further, we showed that the 16 [indiscernible] targets included in the EVX-04 activates human immune cell across different HLA types and that these immune cells can mediate targeted cell cooling, indicating not only in new recognition but also relevant functional impact of these vaccine-induced immune cells.

So collectively, these data highlights EBX 4 potential as a new effective therapeutic cancer vaccines and we look forward to report further data as the program progresses towards regulatory buying later this year.

Another promising program presented at a scientific conference during the summer is our EVX-D1 cytomegaly or CMV vaccine program. So in EVX-D1, we are using AI immunology to design a known target, so optimizing them and also to identify previously unexplored vaccine tires. And at the international Herpesvirus workshop in July, we presented new data demonstrating that [indiscernible] discovered with AI immunology have the potential to control acute infection, latency and also reactivating reactivation in CMV-infected mice.

And this is a key finding as it complements previous results demonstrating the ability of both novel and optimized non-B cell antigens to reduce viral infection. And the data will guide antigen selection for a broadly protective CMV vaccine candid and, as such, represent and a very important step towards for with the EVX-D1 program. So having highlighted progress across our key R&D programs, let's now focus on our AI Immunology platform and the data validating its ability to generate high-quality product candidates.

So AI immunology is clinically validated with positive outcome in 3 out of 3 oncology trials. Preclinically, we demonstrated vaccine proof of concept across multiple disease areas, including cancer, with our art targeting vaccines as well as in taxis diseases with several candidates against bacterial and viral pathogens.

And importantly, the EVX-01 concept is highly scalable with Presento in other solid tumors. And additionally, the novel air-based cancer vaccine concept is used in both our off-the-shelf programs, EVX-04 and EVX-05. So finally, AI analogy supports multiple modalities, including peptides, recombinant proteins, DNA and RNA platforms enabling both pipeline and powering percent. So in conclusion, we've demonstrated strong progress across our R&D pipeline, and we look forward to providing updates as our programs progress.

So with that, I will hand over to Thomas, who will present our quarterly financial results.

Perfect. Thank you, Birgitte. And let me jump straight into the presentation of the financial results for the second quarter of 2026. The main highlights to start with that for the quarter is that we have indeed continued our disciplined resource allocation throughout our strategy direction, of course, and certainly also very much aligned to the priorities around the value drivers that we have defined and also communicated earlier for this year. So full alignment and full progress on those elements.

We are certainly also on track to deliver according to our financial plan. which both shows in the Q2 results, but certainly also confirmed from the cash position that we do have. And the cash position, we can reconfirm, as mentioned by Helen already that we have a cash runway that runs into the second half of 2027. So reconfirmed and maintained from earlier communication also.

Looking a little bit closer to our profit and loss statement for the quarter. Overall, we see a nightly reduced operating expenses mainly driven from our general and administration costs or G&A costs, where we have significantly lower capital market costs in Q2 compared to the same period last year. On the R&D front, expenses do show a slight increase versus last year. but it's fully enhanced with all the progress that Birgitte just mentioned on EVX-01, EVX-04, EVX-05 and again, also those programs are confirmed within our cash runway until the half year 2027.

We reported a net loss for the period of $3.7 million, again, as mentioned already, on plan and following the execution that we've set for this year. Balance sheet, we have a cash position at the end of the quarter of $14 million. We are we are again reconfirming our cash runway and the equity that we also have reflects the result for the first 6 months meaning that we are at $9.5 million at the end of the second quarter, reflecting that versus last year of the net result. So all in all, a good financial performance aligned with expectations and certainly aligned with the progress of our platform and portfolio.

And with that, I hand it back to Helen for some concluding remarks.

Thanks, Thomas, and thanks, Birgitte. So in conclusion, I would want to emphasize that we've seen some really good operational momentum on our asset milestones and actually new program emerging with EVX-05 from all of our activities but still maintaining cash runway into the second half of 2027. We're really excited by the stream of new data that we've continued to generate with the team that continues to validate that AI immunology can deliver products meaningful products for future development.

And that is the core underneath all of our ongoing business development discussions as we continue to process which programs that we bring forward and with which partners. So with that, we are very happy to take questions, and thank you for your attention.

[Operator Instructions]

And this question comes from Thomas Flaten from Lake Street Capital Markets.

Just 2 questions on EVX-05. I was curious if you could perhaps delineate when we might expect to see some more news out of that program. And then if you could elaborate a little bit on the specific role that Duke played in the development up to date?

Sure. Yes. So the collaboration with [indiscernible] has been ongoing for quite some time. They do have a lot of sequencing data from the patients that they're treating in their clinics. So we believe sequencing data for some of those, and we're able to identify. First, we did our presliced approach looking into the profiles of the EV and new antigen expression. And then as EXO we're in parallel progressing and this off-the-shelf concept we're developing. We were able to use some of the same approaches and analyze these samples for identifying conservatives and we were very pleased to see that across these many patients, there were at features indicating that we could definitely generate and off the shelf or the signing of the shelf therapy.

It's still, as I mentioned, a bit early in the development path. We have conducted and concluded we will call target discovery, so selecting the targets that will be included in the vaccine, and we are now heading towards lead selection -- we have designed several different candidates that are now being experimentally tested. And then it's the classical part with R&D-enabling activities and then the first in human study. We have not yet settled entirely on a time line for all of these activities, but that's what we are working on at the moment. So more to come, definitely.

We are now going to move to our next question. And this 1 comes from RK from C. Wainright.

There are a few questions from me, but let me, hopefully, I could go 1 at a time. Starting off on EVX-01. Obviously, it was exciting to see yesterday's news from the Modena collaboration because it validates the program that you have been working on for a while now. So going into ESMO for the 3-year EVX-01 extension data, Birgitte, what would you consider a clinically meaningful durability results that can especially in the stand-alone vaccine period, and how would that help your discussions with either the partners that are currently looking at this program or even the AI model itself that helped generate EVX-01 on a broader perspective?

Yes. So for the EVX-01 clinical data that we would like to see at ESMO is basically that we have almost the same or even improved overall response rate. So we should remember that these patients, advanced melanoma patients, if they only receive checkpoint inhibitors, then almost half of them by the 5-year mark is actually having a severe disease or even, yes, pass away. So there is definitely a high unmet medical need for these patients. So we would like to see that we have durable responses.

So the same number of patients remain in response at the 2-year mark and further that the T cell responses are maintained. So that is -- we would consider that as positive data, positive outcome of this extension base. And then you had an additional comment around how this data would potentially support a partner positive questions. Yes. So there's no doubt that the more data, positive data we can generate would be appreciated by -- in these discussions. And I think the validation that came out yesterday at the personalized cancer vaccine concept, definitely also is supportive in -- or supports us in these discussions.

And we have been waiting the wholesale has been waiting for these Phase III data for a long time. And it's not just a win for the Modena and Merck, but it's actually a win for the whole field. So definitely, we see this as very encouraging and positive and not just -- yes, bad competitor is it's very positive.

Okay. Then going on to the off-the-shelf molecule, EVX-04, in terms of getting it ready to get into the clinic, what are the gating steps here? Is it manufacturing? Is it CMC or making sure that you have enough investigators who would do the right thing when you're starting this into the clinic?

Yes. So EVX-04 we have done target discovery. We have selected the [indiscernible] and now we are conducting IND-enabling activities, so that includes the GMP manufacturing. And then, of course, we need to check that the molecule that is produced is also capable of driving a strong immune response. And then at the same time, we also engaging with the clinical sites, ensuring that we have a setup for testing the EVX-04 molecule. We plan to take this program into the clinic, but we are, of course, always interested and are engaging with companies.

So yes, yes, but it's not necessarily dependent on us entering into a partnership...

And all of those activities are going and on track. So definitely in terms of clinical sites, protocol development, GMP production, compiling the necessary regulatory documentation. So that contributes to our time frame that publicly. So no change or no concern at the moment. We know with all those activities and on track with the communicated time lines of regulatory filings by the end of the year.

I've got a couple more questions. One for Helen. So you -- it's -- you previously even the previous management have been kind of talking about potential partnerships over 2 quarters now. At this point, what can you tell us in terms of where some of these discussions are and if you would like to characterize the stage of the most advanced ones? Where are they at? Are they like the due diligence part, exploratory part or you're almost in the hands of the [indiscernible] and waiting for them to get things put into print?

Sure. That's an obvious -- it's a good question, okay, but 1 I can't really answer as transparent as you would like. I would say in the -- in our oncology conversations, obviously, clinical data that we have that begins talked about, particularly EVX-01 has been very meaningful. But I think the to some extent, the validation of the whole field in terms of having -- seeing a company with a similar sort of program able to bring that forward to a registrational study has quite an impact.

So I think whilst we've been doing various levels of dialogue and diligence, things have been somewhat sort of wait to see how the field pans out. And I think, hence, Birgitte's comments earlier about the positive endorsement of this provides for all of us who have programs that are actually quite differentiated in terms of what they can offer and beyond melanoma as well. So in amongst all of that, I think that the novelty around the IRF platform, the ability to find the conserved antigens from the dark genome has also peaked quite a bit of interest.

Coming in with the second program there in a highly very difficult to treat brain cancer accelerates that interest. So I've been doing BD for 20-odd years and things can go very fast when there's motivation and competition and sometimes it can take 2 years. So I would say that we have active conversations. And obviously, we'll be very happy to update when we can.

One last question from me. So Thomas, when we look at your operations in the first half, the cash use was about $80 million, and it looks like your quarterly burn rate is about like $4-plus million. So against the $14 million that you have in the bank now -- can you walk us through your assumptions of how to get into second half '27? And are you expecting cash infusion either organically or inorganically?

Yes. No, good. Thanks, RK. So maybe the first part of your question. So our cash or cash out is not linear in the sense of each quarter just to extrapolate that. So of course, what we've seen and done in Q2, even in Q1 isn't just automatically to be extracted for the full year. There are some differences. Now we are and will expect to remain on that level that we've communicated also that roughly $14 million for the we might -- and I would expect to be even slightly lower than that. So it's not a round figure as such.

We do have, of course, $14 million, as you rightfully has have seen in -- on the bank account. Please also do remember, of course, that there are some normal fluctuates based on where predominantly DKK based company versus U.S. So there are some fluctuations from a pure ForEx perspective into that also. On top of that, -- so we still do expect that with the runway and with the focus on where we spend, how we spend that we still, as mentioned earlier, can confirm that we are in the second half of 2027.

And we will, of course, utilize the different things that we have available to us. One is also -- not that, that has gone in, I should start paying into the plan in terms of how we communicated half 2 '27, but we do have an ATM facility that we can make use of. And actually, just as of yesterday, we also activated some of that ASM also in the market. So based, of course, on the positive news, as we've seen and the volume in our price.

So we will make use of those type of possibility from an ATM perspective, plus, of course, when we also, at a point in time, announce deals or partnerships that will certainly also add to it. But with the current straight runway and with our prioritized programs, we are very confident that we will go and get into the second half of 2027.

And this question comes from Debanjana Chatterjee from Jones.

This is [indiscernible] on for Debanjana. We had a few questions as well. So the first 1 that we wanted to ask was which glioblastoma patients are most likely to benefit from the EVX-05 cancer vaccine that you are developing?

Specify the specific popular. We are still working on identifying our we're still looking into different patient subsets and looking at the different ERV profiles and seeing what would be the most all set up the most optimal patient population. And further, we are, of course, also looking into standard of care and combination therapies, 1 to be a little bit cautious on combining a vaccine with chemotherapy of the main option of going into those patients that are not benefiting from chemotherapy treatments. But we haven't entirely send on the specifics around the clinical [indiscernible].

Okay. And then as a quick follow-up, so what should we expect as the time line for initiating that first in human clinical trial and what are some key milestones that investors should be watching for before that trial initiates?

Yes. So we are early in the preclinical development. We've concluded on target discovery. So we're using our AI immunology for mining, the patient data and now have a set of optimal that will be included in the EVX-05 vaccine. So we are screening, we have designed several different vaccine candidates are now experimentally assisting those to select the lead candidates. And then it's the classical activities, activities prior to the first-in-human study.

And as mentioned, we are working together with Duke University. We haven't communicated any firm time lines on this program as we need, we need to see, first of all, the selection before we start communicating time lines.

[indiscernible] platform for EVX-04 in terms of delivery methodology, which definitely will we use the expertise and experience there for some of the GMP production side of things. So more to come on the time, but certainly, there's a lot we know about how to bring this kind of platform forward given the way we've done it already for EVX-04.

And then as a final question, so beyond glioblastoma, how broadly applicable do you believe the ERB targeting approach could be across various solid tumors. And then broadly, how does the EVX-05 fit into the long-term strategy of building that AI-driven oncology franchise?

Yes. So we have worked a lot in using immunology to mine patient data across the different indications. And we do see that there are certain patient types where they have a bad antigens. So there's definitely an option of applying this approach more broadly but it's dependent on the profiles of those indications. But definitely more options for scaling this into other solid tumors and also hematologic.

And I think what's interesting is that often where not a high mutational burden. I think that there often is a high frequency, and that's what we've been looking into. So often where there isn't an opportunity to take a personalized approach forward because of low mutational burden, that doesn't seem to be the case with the ERFs more to come on that as we've been teething this part. So we think it really does broaden out the opportunity in terms of what we -- the counter vaccine reach for novel targets.

[Operator Instructions] There seems to be no further questions for today, so I will hand the call back to Helen for closing remarks.

Thank you, and thank you, everyone, for listening in today and for the excellent questions that we've had. We're really excited about the operational momentum that we've been able to deliver but the interest in the programs coming in on the back of a really exciting times for personalized cancer vaccines in the whole field. So exciting things to come and we look forward to updating you further in the second half of the year. Thank you.

Thank you. This concludes today's conference call. Thank you for participating. You may now disconnect.

Good day, and thank you for standing by. Welcome to the Evaxion Business Update and First Quarter 2026 Financial Results Webcast and Conference Call. [Operator Instructions] Please be advised that today's conference is being recorded.

I would now like to hand the conference over to your speaker today, CEO, Helen Tayton-Martin. Please go ahead.

Thank you, and welcome, everyone, to Evaxion's Q1 2026 Business Update Call. I'm very pleased to be joined today by our CSO, Birgitte Rono and COO, recently promoted, we will talk more about that; and Thomas Schmidt, our CFO; and our Head of Investor Relations and Communications, Mads Kronborg.

So if we move to the first slide, just to provide some orientation as to what we will cover today. We will spend a little bit of time on our achievements in the first quarter of this year and some notable changes that we have made in order to address and focus on our strategy. I will then hand over to Birgitte, who will talk through some of the recent highlights from our R&D portfolio and AI-Immunology platform. Birgitte will then hand over to Thomas, who will walk you through our Q1 financial results. And then we will have some concluding remarks before opening up the call for Q&A.

So if I move to the next slide, just to reiterate, we may make some forward-looking statements on the call today, and investors and all listening are guided towards our SEC filed documents.

So if I go past our introduction to Slide 5. I just wanted to, as before, emphasize our 4 key focus areas within the organization and give you a sense of the momentum as we perform an update in each of those areas.

First of all, our core focus around business development and partnering is very much underway and strengthened. By the way, we have reorganized the organization somewhat, and I'll come on to that to focus on the external outreach and positioning of the company to a broader audience and also to raise the awareness of exactly what it is that Evaxion can deliver in terms of products and the platform. And I'm pleased to say that we have many discussions ongoing there, and we hope to report more on that later as the year progresses.

Secondly, in our R&D focus areas, we are delighted to talk about our recent data from our EVX-01 lead program Phase II study in which we were able to update some of the translational data recently at AACR, and Birgitte will talk in more detail about the performance of the cells that we produce in relation to the vaccines given to the patients and the 86% immunogenicity conversion rate we have there. We also were able to present at AACR, a new set of data preclinically in collaboration with our collaborators at Duke University on the scope to use the AI-Immunology platform in glioblastoma. We have always felt that the approach could be applied to other high mutational burden tumors, but also to others where high mutational burden was not a feature, and that is very much part of how we were able to demonstrate the broader applicability of the platform in glioblastoma. And again, Birgitte will speak more to that.

Finally, we were able to confirm the completion of the last patient, last visit in the extension phase of our EVX-01 program and our Phase II trial, and more to come on that later in the year.

More broadly on the AI-Immunology platform, we continued to optimize and strengthen that around its ability to deliver products across our infectious diseases as well as oncology portfolio and again, also in autoimmune disease, again, where we'll update later in the year. But in this first quarter, I'm delighted to say that we were able to show some initial data on a new polio vaccine concept presented in collaboration with The Gates Foundation. And finally, as Thomas will come on to, we have maintained our disciplined allocation of resources aligned to our stated aims with the portfolio and the platform, and our cash runway remains unchanged into the second half of 2027.

Moving to Slide 6. As mentioned, we have reorganized slightly inside the organization. I'm delighted to announce the promotion of Birgitte to the combined role of CSO and COO, which really reflects on how we organize the company and how well it's been run in recent times, but also to enable me, in particular, to have a greater focus externally on behalf of the company in terms of our business development and our investor interactions.

Separately, and in parallel, we were able to welcome Jens Bitsch-Norhave to our Board of Directors. And Jens comes to us with a huge amount of experience in BD and corporate strategy and outreach in general, both from a biotech perspective but more recently from J&J and Hengrui, where he is currently Corporate VP and Global Head of Corporate Development. So we're delighted with the way that we've been able to strengthen the organization to focus on our stated strategy, to build and maintain what we have and build greater partnerships.

So on Slide 7, just to summarize, we remain a lean and capable and focused team in terms of the management organization. Two of the members are here on the call with me today, and Andreas continues to support and drive the organization's innovation strategy around AI-Immunology. And our Board remains the same but with the addition of Jens, as I mentioned.

Finally, moving to Slide 8 to set up our objectives and key milestones for this year. Just a reminder that Evaxion over many years now has the privilege of having a pipeline in Phase II in oncology with our EVX-01 asset in advanced melanoma, our personalized neoantigen-directed peptide-based vaccine, where we've got great data, which Birgitte will touch on in terms of now and what's to come. We have our EVX-03 program, which is a combination of personalized and IRF-based antigens on our DNA platform. And then we also have coming along in preclinical development, aiming for clinical readiness by the end of this year, our off-the-shelf vaccine program, EVX-04 targeted to AML, which will be a single vaccine approach for multiple AML patients. More to come on that.

Infectious diseases, we remain focused on driving forward our preclinical assets, EVX-B1 against pathogen staph aureus, our B2 program against Neisseria gonorrhea and also in collaboration in -- with Afrigen on an RNA platform. EVX-B3, our options partner program continues to move forward with MSD. And before our more recent newer program on Group A Streptococcus is making great strides in initial early discovery component design. And our first viral program is continuing to make progress in terms of confirming the candidate components. So a lot going on in the organization.

In Slide 9, I just wanted to remind the audience of our 2026 milestones and the fact that we have achieved the first one of those in our EVX-01 additional biomarker and immunogenicity data, and we remain on track in terms of updating on the approach of AI-Immunology in autoimmune disease, our 3-year data for the EVX-01 melanoma program, our planned strategy with the EVX-04 AML program and the early work maturing in our preclinical EVX-B4 program against Group A Streptococcus. And fundamentally, we are driving the partnership strategy to focus on the platform and the assets so that we can continue to build value in the company and focus on delivering those into early development where we believe we can add value.

So at this point, I'd like to hand over to Birgitte, who will talk you through our R&D and AI-Immunology update.

Thank you, Helen. So today, I'll be focusing on our lead candidate, EVX-01. And as mentioned, this is our personalized neoantigen cancer vaccine currently in Phase II in advanced melanoma. And then I will present the exciting new data demonstrating the scalability of our AI-Immunology platform into the hard-to-treat deadly brain cancer glioblastoma. And lastly, I will showcase how we have applied AI-Immunology to design optimized vaccine antigens for an improved polio vaccine.

So if you take the next slide. So as Helen mentioned, we presented EVX-01 Phase II biomarker and T-cell immune data at the AACR Annual Meeting here in April. And we reported that 86% of the EVX-01 vaccine target triggered a specific immune response, and this is substantially higher than what has been reported for other similar vaccine candidates. Furthermore, we also reported that 86% of the immunogenic vaccine target induced a de novo T-cell response, meaning that the EVX-01 vaccine specifically triggers novel T-cell responses and not just amplifying existing responses. And this is of great importance as induction of these novel responses have been linked to clinical benefit.

Furthermore, we demonstrated a positive correlation between the predicted quality of the EVX-01 vaccine targets and the magnitude of the T-cell response induced by the vaccine targets. And this high vaccine target success rate, together with this positive correlation demonstrates the strong predictive power of our AI-Immunology platform.

If you take the next slide. So EVX-01 continues to deliver strong data, adding to the already existing and promising clinical and immunological data package. So at ESMO last year, we reported a 75% overall response rate, including 25 complete responders and 92 sustained responses, indicating a durable benefit. So importantly, more than half of the patients converted into an improved clinical response upon EVX-01 treatment. And with the newly presented Phase II immune data, this further strengthens the picture with the 86% immunogenicity and the 86% de novo immune responses, demonstrating broad and consistent immune activation.

So looking ahead, we have a clear development trajectory. We will announce 3-year data, including clinical outcome in the second half of this year. Further, we are evaluating and discussing additional relevant cancer indications and with further trials expected to be conducted in partnerships. And importantly, EVX-01 has already received FDA Fast Track designation, validating both the unmet need and then also the development potential. So overall, this positions EVX-01 very strongly as we move forward into the next phases of value creation.

If you take the next slide. So let's turn our focus to the other promising data set presented at AACR. So in collaboration with Duke University, we demonstrated that our AI-Immunology platform scales beyond melanoma. And here, it's exemplified with glioblastoma or GBM.

So GBM is the most common and the most aggressive primary malignant brain tumor. And despite surgery followed by chemoradiation, outcome remains very poor for these patients with a median overall survival of approximately 15 months and a 5-year survival below 10%. So using our AI-Immunology platform, we have evaluated tumor omics data from 24 GBM patients and demonstrated that a fully personalized vaccine design was feasible for all these cases. And importantly, these designs were based on 2 classes of antigens or classical neoantigens and also antigens derived from the dark genome so-called endogenous retroviruses or ERs. So in 21 out of the 24 designs, they included both types of antigens, 2 vaccine designs included only neoantigens and 1 design relied solely on the ER antigens.

This analysis showcases the flexibility and the scalability of the platform to integrate antigen from different sources, fitting the patient tumor biology. So overall, the data demonstrate that AI-Immunology can address hard-to-treat low mutational burden tumors like GBM and it also supports broader applicability of the platform across different cancers.

If you move on to the next slide. So another example of how AI-Immunology can be used to design improved vaccine was showcased at the World Vaccine Congress. And together with The Gates Foundation, we presented a new polio vaccine concept using AI-Immunology, we designed a novel hybrid capsid antigen and a novel de novo B-cell antigen with the aim of eliciting a strong and broad tumor response against all serotypes. And overall, this highlights the potential of AI-Immunology to reinvent classical vaccines with improved simplicity and also improved breadth.

Take the next slide. So having highlighted progress across the key R&D program, let's step back for a moment and focus on AI-Immunology and the data validating its ability to generate high-quality product candidates. So AI-Immunology is clinically validated with positive outcomes in 3 out of 3 oncology trials. And preclinically, we have demonstrated proof of concept across multiple disease areas, including cancer with our IRF targeting off-the-shelf vaccine concept as well as in infectious diseases with several vaccine candidates targeting multiple bacterial and viral pathogens. And importantly, the EVX-01 concept is highly scalable with potential in other solid tumors beyond melanoma. Additionally, the platform's applicability in challenging cancer indications was further validated in GBM. So finally, AI-Immunology supports multiple modalities, including peptides, proteins and DNA and RNA, enabling both pipeline and also partnership potential.

So in conclusion, we have demonstrated strong progress across our platform and our R&D pipeline, and we are looking forward to keeping you updated as we advance our programs further.

So with that, I will now hand over to Thomas, who will present our quarterly financial results.

Yes. Thank you, Birgitte. And as mentioned, I will now then present and take you through our Q1 '26 results. The highlights of the first quarter of the year really is a continued discipline that we have applied in our resource allocation, of course, aligned with our strategy and certainly investing into our value drivers. So really according to plan. And that also means that we are on track to deliver what we expect of an operational cash burn of roughly USD 14 million for 2026. That also underlines and reconfirms that our cash runway is into the second half of 2027 and remains as such. Also, as earlier communicated, not assuming any partnerships or deals that we will hopefully be making and communicating within that time frame.

Looking at the P&L, we have operating expenses overall more or less in line with last year, but slightly reduced. It comes from our R&D with a minor increase as we continue, as mentioned before, to progress and advance our pipeline and programs according to plan. On the other side, our G&A expenses are slightly lower versus last year, also mainly driven by the fact that we have lower capital market costs in Q1 '26 versus the same period in '25. The first quarter resulted in a net loss of USD 3.6 million, again, according to our plan.

On the balance sheet side of things on the next slide, reconfirming once again, our cash position and equivalent end of the quarter stands at $18.4 million, which confirms runway until the second half of '27. And the total equity has been reduced since year-end, really as a result of the net result of the first quarter, meaning that we have USD 13.2 million as equity at the end of the quarter. So all in all, financials according to plan, allocation into our main priorities and cash runway confirmed until the second half of 2027.

With that, I hand it back to Helen for some concluding remarks.

Thanks, Thomas, and thanks, Birgitte. And so I would just like to emphasize that we believe we've made a great start to 2026, achieving the first of our milestones with a really encouraging translational data from EVX-01. We've got various presentations that have been made that validate the capabilities and scalability of the platform, as Birgitte has explained.

Business development remains a key priority in terms of engaging with organizations on the value of the assets that we have and the capability to develop those assets as we've talked a bit about. And the cash runway is maintained through to the second half of 2027. So we are rigorously following execution of our strategy and engagement externally and making great progress.

So with that, I would like to hand back over to take some questions by the operator.

Our first question comes from the line of Thomas Flaten from Lake Street Capital Markets.

Two for me. With respect to the 3-year EVX-01 data, ASCO is obviously too soon. But should we anticipate something like an ESMO readout? Or will you do it independent of a broader scientific meeting?

So we will be updating in the context of a scientific meeting. We will not be sort of outside of that, that's not our intention. And we'll confirm which of the 4 conferences it will be once we're able to -- once abstracts are released.

I think the GBM data that you put out, albeit early, was very exciting, and obviously, a disease state and great need. Is it your strategic intent to take that into humans? Or would you seek a partnership based on the data you have now and perhaps some additional preclinical data?

So we are very excited about the data. We agree it's really interesting and it's really exciting in a very difficult-to-treat disease. We would anticipate that that will be something that we will be partnering. It sort of strengthens the overall personalized approach that we have developed with EVX-01, but probably more to come on that as more data and discussions mature, but it would be a partnering approach for that one, too.

Our next question comes from the line of Michael Okunewitch from Maxim Group.

Congrats on all the great progress. I guess to kick things off, I'd like to ask just a little bit about expansion and I guess, your design philosophy and strategy around that. So first off, when thinking about targets for expanded indications in cancer, in particular, is the plan to go after other diseases where PD-1s have historically been ineffective due to that synergistic activity of directing the antitumor immune response?

So I think we've taken a lot of parameters into account. But Birgitte, do you want to comment on how we have been marshaling the approach internally to focus on the rare diseases?

Yes. So as mentioned, we are looking at multiple different antigen sources currently, and there's further development in this area in the company. So we would like to be able to provide a cancer vaccine for all patients independently of their antigen profiles or landscapes. So we have so far looked at more than 30 different indications, mapping out their seasonal burden, their ERV burden, et cetera, and can see that for many of these indications, we're able to -- with the capabilities we have currently to design a high-quality vaccine. And of course, one would need to further dive into medical need and current treatment landscapes to find the optimal subpopulations where our therapies would fit, but not necessarily in PD-1 low patient, it could also be in high. So it's mostly -- we are mostly focusing on understanding the antigenic landscape and fitting our therapies towards these profiles.

When thinking about designing new vaccines, do you find that it makes more sense to use one personal vaccine and then see if you could expand that to multiple tumor types with the same vaccine for more universal coverage? Or does it make more sense to go tumor by tumor and create a new back of targets that are directed specifically at the common target for that given tumor type like melanoma or like glioblastoma and have an individual vaccine candidate for each of those different cancers?

So the way that we are approaching this is to look into a lot of data from certain indication and understanding, as mentioned, the landscape. If we do see that there are these conserved antigens, so antigens that are shared across patients, we would definitely develop an off-the-shelf vaccine just due to the fact that the statistics are more simple and also the cost for the manufacturing would be way lower than for a personalized approach. Further on, you can -- if there's an off-the-shelf therapy, you can immediately treat the patients and not have to wait for that personalized batch to be ready.

So that's -- everything comes back to the patient omics data and the profiles that we are seeing in our analysis. For some indications, we know that developing an off-the-shelf cancer vaccine would be very challenging. So it clearly depends on these different biological profiles.

I think the EVX-04 illustrates just where in that setting, I think, the high level of conserved has enabled us to produce a single vaccine for those patients.

I appreciate the additional color and looking forward to the 3-year data coming up later this year.

We will now take our next question. And this question comes from the line of Danya Ben-Hail from Jones.

Congratulations on the update. You mentioned that there are several parallel partnerships and discussions. Can you provide more detail on whether these discussions lean toward broad platform licensing or specific asset-based collaboration in future?

So we obviously can't say much at this point. I think we have stated the priority around partnership on EVX-01. But as you've heard, that has broader applicability than melanoma in our minds. And that has obviously also gathered interest externally with partnering conversations also.

Across our infectious diseases portfolio there are a number of assets there, which are of interest to a number of companies. So we can't really provide any more details than that. Suffice to say that we are trying to be strategic around the way we have the partnering discussions in terms of maximizing the value, whether it's from an asset group in infectious diseases or the approach with something like the personalized EVX-01, EVX-03 cancer vaccines. So obviously, we will -- as soon as we can tell you more, we'll be delighted to do so, but we're pushing forward on a more strategic basis, if you will, around how to get the most value out of the assets that we can produce from AI-Immunology.

Just one more question on the autoimmune platform part. So we should expect more details in the second half?

Yes, that's our current plan and aligns to -- as is always generally with Evaxion, generally aligns to scientific relevant conferences to report on data.

There are no further questions for today. I will now hand the call back to Helen Tayton-Martin for closing remarks.

Thank you. Thank you very much. And thank you to all those who listened into the call, and thank you very much for the questions that we received. I think in summarizing, we are very enthusiastic and excited about the performance so far in Q1 2026. We are really just getting started and we are achieving our milestones as we have stated them to be. So very excited about the initial data, very excited about the additional updates to come later this year.

With that, I'd like to thank you very much, and I think we'll be closing the call.

Thank you. This concludes today's conference call. Thank you for participating. You may now disconnect.

Good day, and thank you for standing by. Welcome to the Evaxion Business Update and Full Year 2025 Financial Results Webcast and Conference Call. [Operator Instructions].

Please be advised that today's conference is being recorded. I would now like to hand the conference over to your speaker today, Helen Tayton-Martin, CEO. Please go ahead.

Thank you, and good morning, everyone. Thank you for joining us for Evaxion's business update following the reporting of our 2025 full year financial results yesterday. And apologies that this call is 24 hours later than we anticipated for technical reasons, we are delighted to be here today. My name is Helen Tayton-Martin, and I am honored to be leading this call for the first time as Evaxion's CEO.

We move to the first slide. Okay. So on today's call, we will review the achievements of 2025 and touch on the milestones we anticipate for 2026. Our Chief Scientific Officer, Birgitte Rono, will then walk through our key R&D updates for the year, including the latest innovations from our AI immunology platform, after which our CFO, Tom Schmidt, will walk through our 2025 financial results before we close with a few concluding remarks and take questions.

Right. Moving to the next slide. And of course, our comments and presentation today may contain forward-looking statements and all references on today's call, I'll refer to our filed SEC statements and specifically, our most recent 20th annual report for 2025 -- 2025 filed yesterday.

So moving to the next slide. I will start with our 2025 achievements and our 2026 milestones. In 2025, we were very pleased to report tremendous progress across all pillars of the company. First of all, in business development, we were delighted with the progress in our collaboration with MSD and our infectious disease portfolio, with the decision by Merck to exercise its option over our EVX-B3 program candidate. Whilst the target for this program is not disclosed, we are very proud that this represents the first in-licensing to our knowledge of an infectious disease vaccine candidate identified and validated to an AI discovery platform.

Whilst MSD chose not to exercise its option over our EVX-B2 candidate in gonorrhea, we remain very excited about the data and the prospects for this program, over which we have retained full rights and have seen significant interest. We were also pleased to enter into a collaboration with the Gates Foundation on the design of a new polio vaccine and are also seeing significant interest in our platform and pipeline programs more broadly from a number of parties.

In R&D, we were very pleased to be able to present very positive [ to ] Phase II data at ESMO on our EVX-01 program with a personalized neoantigen-directed cancer vaccine in advanced melanoma patients. We also presented preclinical data at ASH on our first cancer vaccine we shared [ at antigen directed ] to a conserved endogenous retroviral EBR elements that we have identified in AML patients with our EVX-04 program.

In our infectious disease portfolio, we were also able to move forward a new program with candidates identified from our AI immunology platform against [ Group A strep Coke ]. On the platform itself, the team has continued to innovate and use platform to not only identify optimal vaccine candidates, but improve their design biology for product delivery for us in our new automated module. And Birgitte will touch on all of these achievements shortly.

We were also honored by the recognition of our AI immunology platform by the Galien Foundation for AI advances in human health. And finally, we were very pleased to see the capital influx of the business last year through financing, business development and the use of our ATM, which now gives us action a cash runway to the second half of 2027. And Thomas will talk more to this later.

So moving to the next slide. Just as a reminder, our action has built a broad novel product basis pipeline of assets from its unique AI immunology platform. clinically validated with the cancer vaccine space with our EVX-01 [ peptide ] base vaccine in advanced melanoma that's supported by assets and data on DNA and RNA platforms and together with a preclinical pipeline of infectious disease vaccine candidate, focused on challenging targets remaining intractable with conventional approaches and subject to significant medical need.

On to the next slide. This unique capability with AI immunology is something that we have also begun to investigate within the autoimmune field, given a wider range of diseases driven by autoimmune attack and the direct applicability of our platform to focus on immune mechanisms in disease. Autoimmune diseases affected over 14 million patients annually in the U.S. and are characterized by chronic debilitating conditions with treatment options focused primarily on the symptoms rather than the underlying cause of disease.

Moving on to the next slide. This is why we believe our AI immunology platform is strongly positioned to focus on underlying disease mechanisms with greater specificity to identify autoimmune disease targets, which can be approached in different ways. There will be more to come on this later in the year. So finally, in the next slide, turning to our 2026 milestones. This year, we will be updating on our EVX-01 program with additional biomarkers and immunogenicity data, AACR and then the clinical data, 3-year data towards the [ later ] towards the end of the year.

Well, we will be talking more about the autoimmune applications of our AI immunology platform and bringing forward data on our new EVX-B4 candidate in Group B [ rubric ] in the second half of the year. And finally, be ready to submit a regulatory application for our next EVX-04 candidate vaccine candidate for the shared her antigens in AML by the end of the year. And throughout, we remain committed to driving value from both our platform and our pipeline assets to partnership for our shareholders and patients.

I'll now hand over to Birgitte to update you further on our R&D achievements.

Thank you, Helen. So 2025 marked a turning point with significant advancement across our R&D pipeline and also and our AI platform. And additionally, as Helen alluded to, we also entered into the in-licensing agreement with MSD on the EVX-03 program. So our 2025 focus has been on strengthening our platforms predictive power, maturing key R&D assets and are building the foundation for future partnerships. So the 2025 achievements position us well as we move towards the data with milestones in 2026, that Helen just presented. So with that, I will begin by walking through individual key programs and platform development.

So next slide, please. [ EVX-01 ], our personalized peptide-based cancer vaccine in advanced melanoma continues to deliver strong clinical data. So our 2-year Phase II data presented at an oral session at ESMO in October showed strong clinical outcomes, including a high objective response rate of 75% and complete response rate of 25%. Notably, 92% of the responders remained in response at this 2-year mark.

Key biomarker data included the very high [ immunogenicities ] rate with 81% of all the individual new antigen administered across patients, giving rise to a specific T-cell response. So this very impressive heat rate outcompete data from similar programs conducted by others. And this truly underlines the precision of our AI immunology platform, to identify better than vaccine targets.

Two key milestones are expected for this program, as Helen also alluded to, additional biomarker and [ genicity ] data expected in the first half of '26, and we also plan to communicate the 3-year data from a subset of patients that are currently in expansion part of the Phase II study, and that will be reported in the second half of '26. So importantly, we aim to conduct future trials in partnership ensuring the broadest possible impacts for patients.

So moving to the next slide and EVX-04, our after-shelf therapeutic vaccine for acute myeloid leukemia or e-mail, we have generated a compelling preclinical base evidence supporting its development. In this program, we are focusing on a completely novel class of tumor antigens, so-called endogenous retroviruses or ERVs that are selectively and highly expressed in AML blast, making them attractive as therapeutic targets. So with AI immunology, we have identified millions of shortages from patient sequencing tumor data and designed the [ EVX-04 ] vaccines with 16 optimal ERV [ anti fragment ] selected based on craft patients [ pellets ] and also on the immunological potential.

So key data include invite vaccination studies, demonstrating that all of these 16 fragments in the vaccine induced a strong specific immune response and further that EVX-04 prevents tumor growth in several of our tumor [ virus ] and induce strong T-cell responses. So again, these findings reinforces the power of our platform. And here, we have expanded it to uncover unique tumor antigens that are not accessible through traditional discovery methods.

Next slide, please. As we progress towards clinical business for EVX-04, we have completed key steps, including antigen selection and lead development we have conducted preclinical efficacy studies and are currently conducted further human cell-based translational assays. CMC work and GMP manufacturing are advancing according to plan. And the next major milestone for this program is the submission of the clinical trial application in the second half of '26, which enabled first in human system. So this program is a prime example of how AI immunology accelerates vaccine design from concept to clinic.

So next slide, please. Now turning to our key indexes disease programs. So after retaining the full global rights to EVX-B2 late last year, we are now fully in control of the development of this highly differentiated vaccine candidate targeting -- gonorrhea. So our preclinical data package is strong and comprehensive demonstrating significant protection in a mouse infectious model. We have demonstrated broad efficacy against 50 clinically relevant -- dates reflecting coverage across diverse strengths and further induction of significant [ una ] and cellular responses in mice, and we have also demonstrated a well-established mechanism of actions supported by potent antibody-dependent complement-mediated killing.

So collectively, these results position EVX-B2 as one of the most advanced and differentiated infectious disease preclinical gonorrhea vaccine candidates in an area of high unmet need where no approved vaccine exists today. So given the strength of our data, we see a clear opportunity to engage with potential partners to progress the program towards clinical development. So next slide, please.

So a number of our key infectious disease vaccine program is EVX-B1. In this program, we are developing a margin target vaccine against cytomegalovirus or CMV and instead of relying on a single glycoprotein or limited set of glycoproteins, the program integrates both these well-described glycoprotein and novel antigens to target the prior -- from multiple complementary angles.

So this broad multicomponent strategy is designed to enhanced vaccine efficacy and also to reduce the risk of viral escape. So we have applied AI immunology for both antigen optimization of the known glycoproteins and for identification of 2 novel antigens.

So first, we improved these established CMV antigens that are essential for virus neutralization. And as part of this, we have engineered the glycoprotein B antigen, by locking in a prefusion state. And this AI analogy designed a construct has demonstrated a superior neutralization capacity compared to the native program.

And secondly, we are identifying and validating entirely novel antigens and several of these -- they have already demonstrated the ability to inhibit [ Vinten ] further, we are characterizing them at the moment. So supported by this strong preclinical data, EVX-B1 represents a highly promising program for continued development and for future partnership discussions.

Next slide, please. So now turning to the recent development of our AI-Immunology platform. So our AI-Immunology platform continues to expand capability. So the platform integrates [ multiomic ] data sets to generate ranked antigen lifts within 24 hours. So in October last year, we launched a an automated vaccine design module enabling sequence and structural optimization directly from this short-listed engines.

At this end-to-end automation significantly reduced cost development time and also this. So next slide, please. So more specifically, the automated module enhances design of [ Sage ] antigen constructs, enabling higher expression, better formulation and improved manufacturability. So this capability directs the design of [ salable ] antigen constructs and also stabilizing antigens using in various posing producing more reliable antigen construct vendor, wire side variance. There is a faster and more cost-effective design cycle fully integrated into our antigen discovery and vaccine optimization workflow. So this strengthened the foundation for all of our programs across oncology and infectious diseases.

So in conclusion, we have seen strong progress across our platform and our R&D pipeline, and we are encouraged by the momentum and we look forward to keeping you updated as we advanced to 2026.

And with that, I will now hand over to Thomas, who will go us through our financial business.

Yes. Thank you, Birgitte. And also a warm welcome from my side to our call today. And I will now walk you through the financial results for 2025. So turning to the next page. We have, throughout 2025, been really successful in expanding on our cash runway and also strengthening on our equity side. This has happened throughout the year through public offering and the use of ATM we did in January, followed by the MSD exercise fee and the ATM used in September. And furthermore, the exercise of investor warrants from our January offering in October and November, all summing up to a cash inflow of USD 32 million.

Furthermore, as also shown on this slide, our EIB debt-to-equity conversion done in July of USD 4.1 [ billion ]. We've reduced certainly our cash -- future cash out and thereby certainly also expanding and extending our cash runway. And finally, with our filing in December of our prospectus supplement regarding our ATM. It has now created us with further flexibility ability and options as we move forward with expanding our pipeline and platform also. So really, really underlines the strong execution throughout the year.

And turning to the next slide. that also leads into the highlights of 2025, where we really have delivered on all the targets that we set and we are progressing towards our aim of becoming a sustainable self-funding business. Both revenue and costs have improved while at the same time, we are continuing to invest in our platform and in our pipeline programs. As just mentioned on the previous slide, activities and execution of the MSD deal, the EIB debt conversion, our ATM and capital market activities have not only improved our cash position and runway, but has also significantly strengthened our equity. And with improved cash runway and equity -- equity we have created more stability and certainly have also reduced uncertainty. So I think that is really, really also a highlight for '25.

And again, with the update of [ F3 ] and ATM, we have removed the constraints of baby shelf and also provides us far better flexibility and options in support of our long-term strategic initiatives and also the long-term plans we do have.

Next slide. is on our profit and loss statement. As I just mentioned, revenue has improved, but also we've improved on our operational costs. So we've actually been successful in long in our operational spend whilst at the same time delivering on the quality that we would want to do from a pipeline and platform perspective. revenue certainly stems from our NST option exercises, but also important to mention, we also had a grant from the Gates Foundation that also has come in 2025 -- apologies. Net financial position of [ $4.6 million ] is driven by a premium that we received from -- our debt conversion -- debt-to-equity conversion from [ EIB ] and against that goes remeasurement of a derivative liability as some of our warrants or our [ launch ] from the public offering in January were in a different exchange setting, so the USD versus our reporting of DKK.

Net loss for the year, [ $7.7 million ], certainly a better in compared to last year. And as I said before, also a good step on the way of becoming a sales funding and profitable business. Next slide, on the balance sheet. since we ended the year with a cash position of USD 23 million with a runway that now is extended into half year 2 of 2027 in certainly also a significant improvement compared to last year. And this, of course, will be used for operations expenses and investing into our platform and pipeline.

We currently have an outstanding -- we have a total outstanding ADS of $8.3 million when assuming that all shares have been converted into ADSs. We've also, through the investor warrants exercise has been reducing the outstanding warrants in terms of ADSs by $1 million. which leaves another [ 2.8 million warrants ] outstanding. So also an improvement in that and really drives in the right direction.

So in summary, from a financial position, we have during 2025, established a far better foundation that really makes us puts us in a good position to continue our execution of strategy and business for '26 and the years beyond.

With that, I hand it back to Helen for some final concluding remarks.

Thanks, Thomas. And just moving to the last slide. In summary, 2025 was a year of strong operational momentum for Evaxion, in which we achieved several key milestones. Overall, we strengthened the business considerably to the validation of our strategy with our AI-Immunology platform, delivering on both data and partnerships. This, in turn, has enabled us to both strengthen our financial position and consolidate our position as a leader in AI-based of discovery, design and early development. With a number of potential partnership discussions ongoing, we are already funded into the second half of 2027 through the financial milestones achieved in 2025. So we're in a good position to move forward through 2026.

With that, I'll hand over to the operator for questions.

[Operator Instructions]. Our first question today comes from the line of Thomas Flaten from Lake Street Capital Markets.

Maybe to start broadly, Helen, you've been in the seat now for a few months. I'm just curious if you could provide some overarching commentary on what you have implemented or are going to implement -- any changes in strategy? And any bigger picture notes like that, that could help us with the context of your tenure?

Sure. Thanks. Thanks for the question. Yes, I joined at the end of November last year. So the last 3 months have flown by. But I already have a strong impression from my prior seat on the Evaxion Board. In terms of bigger picture, changes. I think the fundamental action remains really strong. And in fact, I think they have only got stronger through 2025. So the ability to have an AI platform that is built up over many years, many iterations, grounded in data and testing for that data in the lab, and ultimately in the clinic has really strengthened the core offering. So I remain really excited about the power of the platform in the oncology space and also in the infectious disease space. And I think we're sort of seeing a lot more traction around what we can do with the platform now from external engagement.

So I think the fundamental strengths and core of what Evaxion has to offer is even stronger now than potentially before. And I think in a world of AI, everything, actually getting to products, actually producing candidates that can generate vaccines that generate a biological response and the clinical response is meaningful and is becoming recognized as meaningful, certainly in our partnering conversations, et cetera. So I think that, that is core.

So clear observation I had before coming into the company and certainly strengthened by all my observations within it. and even more impressed by the team that's in place that can deliver on this. I think in terms of the overall strategy, what have Evaxion has done well is that early discovery, the early validation, that deep scientific and informatic embedded expertise, and we can certainly bring things forward into early clinical development, late preclinical, early clinical. And I think what we're going through at the moment is a process of really optimizing where we see the most value in the near term, both in terms of our oncology assets, but also within the infectious disease area. I think we are not positioned to take too much further forward into the clinic. So we've been very cautious about that but we certainly see strength in getting interest from external parties around the assets that we've already got. And actually, the capability of the platform. So there's not a fundamental change to strategy, but I think a sharpening and the deepening of focus around the assets that will have the most value. I hope that's helpful.

Yes, that's great. And just keying off of your last comment there about taking products into the clinic. You mentioned with EVX-04 in Birgitte's presentation that you would be looking to submit regulatory paperwork. Is that a product that you think you have to take into Phase I given that herbs are a bit new, a bit different in order to attract a partner interest?

I think that's a very good question. We are certainly preparing to take it into the clinic, and we believe that we can do that to gain some initial proof of concept. There's a lot of interest around the platform at that particular metacandidate antigens in the vaccine. I think we're doing some further validation, which I think will continue to strengthen it.

So the answer in short is not necessarily, but clearly, the more critical validating data that we can add to the package. The stronger the value proposition to an external partner, and that's obviously what we're all about is maintaining the -- building the value for as long as we can to strengthen our position. And I think we're very confident about what we can do with it preclinically and potentially clinically.

Our next question today comes from the line of Michael Okunewitch from Maxim Group.

Congrats on all the great progress you made. I guess to start off, I'd just like to see if you could comment a little bit on the partnering efforts for EVX-01. And in particular, if there's anything that you've heard either in your feedback from partners that you think you're still would be particularly important for us to watch for from the upcoming data releases, whether that's the 3-year data or the biomarker immunogenicity. Is there anything in particular you think is key for driving these partnering discussions?

So that's a really good question. And I mean, clearly, the cancer vaccine space has had something of a checker passed -- way back, but more renaissance, I think, in the checkpoint era. And I think our data is certainly resonating with companies who are interested in the cancer vaccine area, understand the nuances around getting, I think, strong cancer -- antigens, [ presliced ] cancer antigens for not just immune recognition but for clinical benefit.

So the -- it is a complex therapy to administer, but it is also potentially an effective therapy. And I think the sorts of things that gain interest of the -- not just the response rate that we've seen in 2 years, 1 year than 2 years at ESMO, but also the recognition of the antigens, the numbers that the [indiscernible], and I'll ask you to add comment to this as well.

So I think the -- we're in a strong position with that updated clinical data package that we have, the translational data, I think it's going to be interesting. It continues to -- so why and how the immune response is happening in parallel to the clinical response. So that is, I think, a differentiator and also in the population, the advanced population rather than adjuvant melanoma population.

And clearly, I think we're also seeing interest in this whole approach in other high mutational burden cancers too. So beyond melanoma knows of it. I think those are the differentiators thinking about where else this is applicable accolading the different biological parameters, the translational insights that we're seeing that's somewhat different to how others have reported on this with similar approaches. So quite a bit of interest. I think the number -- to be honest, a number of companies are on the fence, but we're looking with interest and very interested in the shared approach -- the share approach that our EVX-04 program offered.

Birgitte, do you want to add some further comments?

So there's no doubt that the ability of our AI immunology platform to identify the relevant targets and is getting a lot of interest from potential partners and also from the academic community. And with this 81% hit rate, as we call it, I think this is very impressive. We have, of course, looked at other similar programs and seen that most of them are reporting hit rates way below 60%, meaning that the antigens that they are including in their vaccines are not all able to induce a specific T cell response. And this is, of course, a testament to the position of our platform. So that's one of the key elements.

And another point that I would like to make is that we do see EVX-01 as not just a therapy for advanced melanoma. We believe that the same concept can be very useful in other occasions where there are a high mutational burden, meaning that there are several antigens to choose from. That includes many of the high prevalent cancer indications. It could be non-small cell lung cancer and also some of the colorectal cancers.

Thank you. I appreciate that additional color on that. And then as a follow-up, I wanted to ask if you could provide a bit more color on how you're applying the AI immunology platform to autoimmune disease. You identified this as a new area of interest. And do you expect that this would be more focused on allergies? Or would you focus more on the major large autoimmune and inflammatory diseases. Any additional color you could provide on that would be helpful?

Sure. I think the first thing to say is it's early in terms of our prioritization of the indications, but we've certainly done some work around that based on parameters, which I'll -- Birgitte you're happy to comment on that, I think, high level in terms of what's guiding where we focus will be -- that would be good.

Yes. So we have done a lot of analysis on most prevalent autoimmune diseases, and we do see a clear fit for our platform. I mean, we, of course, need to further improve it and build a few additional smaller unit that allows us to apply the immunology. But we do have many things in place that can be directly applied in this area. So we will, of course, share more when we have done both analysis on which key indications we will pursue and also when we have done a little bit more work on adjusting AI-immunology, so it fits these diseases. But we should remember to say that there's a lot of these smaller units we call them building blocks that we can directly apply for these tax of diseases. So not only for autoimmune diseases but also for other diseases where there is a strong immunological component. So of course, we need to build a little bit, but the majority is already in AI-Immunology.

Our next question today will come from the line of RK from H.C. Wainwright.

Thank you. This is RK from H.C. Wainwright. Just to start off, Helen, a quick question for you. You have basically, we've been an architect and multibillion dollar balances at that [ immune ] especially the large deal that was transacted with GSK. And also, you have heard a lot of experience in transactions. And while Evaxion is technically a very strong company, they have always had a difficulty in translating that language into meaningful transactions. Of course, Merck is a pretty strong partner.

Based on your experiences, and how you manage to translate that. What sort of discussions could you have at this point? especially when talking with large-cap pharma, I'm convinced them that an AI tools predictability is as good as a physical assay and get them to start looking into some of the products that Evaxion is generating.

Thanks for the question. I think there are probably multiple dimensions to answer that question to the extent that it is possible to answer it at this point. One is that A lot of this is to do with timing. It's to do with data that validates that it's more than a sort of an AI platform. And I think actually, the fact that we have the scope to validate and iterate candidates target discovery with candidate development with cancer validation is something really novel that we're generally out there. And when I said timing -- there are obviously many of the large pharma, most of them will all have in-house AI platforms running in one form or another.

But I don't think there are many that have got this sort of integrated long sort of longitudinal depth of expertise that actually has. So really, is that this is about crystallizing the offering through the validation of the cannabis we have and sort of being in dialogue with the right people. And you mentioned my background was a long time, 17 years in my precise company, building relationships establishing contact, understanding and listening to strategy, looking at the wider picture. These are all things that are very much part of how deals get done. And ultimately, it's down to relationships and credibility and really having something that fits the need.

And all I can say at this point is we're reworking up some of those approaches and some of those themes in terms of how we are approaching potential partnering interaction, I have to say that the action team is well known with quite a few of these groups, but we're building and expanding that profile. And I think that's critical to the future success in partnering conversations. So it's being what you say you are in front of the right...

Perfect. Then going into relationships with Merck, especially regarding EVX-B2, Merck decided to extend the evaluation of the molecule rather than exercise the option at this point. Is this a function of them trying to do additional experiments or functional assets? Or are they requesting from new additional work so that they can come to a conclusion?

Sorry, are you referring to the extension that they had last year before the opt decision there?

Yes, yes.

I mean we can't really comment on, obviously, the combination nature of the interactions. All I can say is that sometimes is sort of R&D programs when they are back and forth and shared between organizations don't always run to plan. And so sometimes that requires looking at things again. But ultimately, then there are time frames around things, which have to follow through. So I think there are reasons for not taking sort of obviously, their reasons not multidimensional.

All I can say is that we remain really excited about the data. We actually continue to build data on the program internally throughout that period of time as well. So we feel very, very bullish and strong about the data package but how and why I wanted to do that work? Or is it something we probably we can't really add any more commentary on.

Okay. On the EVX-01 durability, Birgitte, so you have shown 92% of the responders showing sustainability, be it 24 months. As we are looking forward to the 3-year durability what sort of exhaustion markers are you going to be tracking so that we understand how well the durability is.

Yes. Thank you for that question,. So it's correct that 92% of the responders remained in response with this 2 year mark. And I guess your question was related to the T-cell extortion -- as yes, we do a deep scar profiling, looking at activation marker, extortion markers and also at different phenotypes of the T-cells, so including CD4, CD8, but also looking into whether there are regulatory T-cells coming up. And so far, we have demonstrated that -- the profile of the T-cells are very favorable. So in more of the activation or effective type of sales and not too many that are having exhaustion markers.

We also see that there's a like dominance of CD4 T-cells and with some patients are also mounting a CD8 T-cell by time. So we have -- during this extension phase of the study, we have been collecting additional blood samples that are currently being analyzed in our lab. So not to comment on that, but it's very exciting. And since EVX-01 is giving us a immunotherapy in this extension phase, we're also very curious of understanding what EVX-01 can drive on its own without having the background of the checkpoint inhibitors.

Okay. One last question from me. This is on the EVX-04...

In the interest of time, we will move to our next question. And our next question comes from the line of Daniel Ben Hill from Jones.

On the autoimmune disease program, can you provide more detail on your strategy for validating early candidates?

Thanks for the question. I mean it's early, and we probably cannot provide more details. But, Birgitte, do you want to comment on how we think about it.

Yes. So the first step is to settle on an indication. So we have done landscaping. We've done dianalysis on looking at the top 10 most prevalent ultimate diseases, and we are now narrowing down which one could be the most, I would say, interesting from a -- from our perspective, where there is a nice fit for AI-Immunology. And so that work is ongoing. We've almost completed it. And next step is to focused on building the additional smaller units that we will be needing in AI immunology to enable us to develop therapies for these diseases. And in parallel, we are also sitting on mouse models in our lab, so ensuring that we can also test the candidates that AI-Immunology is designing. So that is the current plan. So pretty traditional way of analyzing our existing the candidates that AI immunology is designing.

Thank you. This concludes today's question-and-answer session. I will now hand the call back to Helen Tayton-Martin, CEO, for closing remarks.

Thank you very much for everyone participating on the call today. It's been a great year of 2025 of transforming the company for Evaxion delivering on multiple milestones, leaving us in a stronger financial position than for some time, where we hope we can take the company forward and deliver on our 2026 milestones and continue to strengthen the value that comes from the platform and the asset.

So thank you for your questions and your engagement, and we look forward to our next update. Thank you. Bye-bye.

Thank you. This concludes today's conference call. Thank you for participating. You may now disconnect.

Hello, and welcome to the Evaxion Business Update and Third Quarter 2025 Financial Results. [Operator Instructions] Please be advised that today's conference is being recorded.

I would now like to hand over to Birgitte Rono, Interim CEO and CSO. Please go ahead.

Thank you. Good morning and good afternoon, and thank you all for joining our Q3 2025 business update and financial results conference call. I'm Birgitte Rono, Chief Scientific Officer and Interim CEO of Evaxion. I'm joined today by Thomas Schmidt, Chief Financial Officer; and Mads Kronborg, Vice President of Investor Relations and Communications.

I'll begin by walking through the agenda for today's presentation. I'll start with a brief introduction, followed by an R&D update, and then Thomas will present the Q3 financial results. And lastly, after a few conclusive remarks, we will open for questions.

I'd like to remind everyone that today's presentation may contain forward-looking statements, and these are subject to risks and uncertainties, and actual results may differ materially.

First and foremost, I'm pleased to welcome Dr. Helen Tayton-Martin as the new CEO of Evaxion effective November 24. So Dr. Tayton-Martin brings extensive biotech leadership, fundraising and partnership experience. Helen co-founded Adaptimmune and has held several senior executive roles in Adaptimmune. Helen holds a PhD in molecular immunology and MBA and with more than 30 years of experience in early research through project approval, Helen is an ideal candidate to lead the next stages of Evaxion strategy. This also means that I will return to my previous role as CSO and with Helen's transition from Director in the Board to CEO, Jens Bitsch will join the Board as an adviser and observer with the intention to seek election at the next AGM.

Since the last business update, we have made several significant achievements. Historical in-licensing of EVX-B3 by MSD, providing significant cash and validation. The extension evaluation period or the evaluation period for EVX-B2 has been extended. We have several ongoing partnership discussions, though market uncertainty affects deal climate. We have presented two-year clinical efficacy data for EVX-01 at the ESMO Congress, and we have added EVX-04, a novel therapeutic cancer vaccine for acute myeloid leukemia to our pipeline. Further, we have expanded our AI-Immunology platform with an automated vaccine design module, improving quality and reducing vaccine design time. Lastly, we have strengthened our financial position, and we now have a cash runway extended to second half of 2027. And this is based on a USD 7.5 million option exercise fee received from MSD and an additional capital market funding sources, and Thomas will share detail around this.

So, as mentioned, one of the main highlights of the quarter is the MSD transformative deal. So, in September 25, was a historical moment for a vaccine with MSD or Merck exercising their option on EVX-B3. So this was the first ever in-licensing of an AI discovered vaccine candidate by a major pharma company. And as mentioned, the $7.5 million exercise fee extends our cash runway significantly.

The deal confirms our strategy of value creation through partnerships even with industry giants like MSD. It also validates our AI-Immunology and R&D pipeline and further ensures the development of EVX-B3 without cost for Evaxion. And not related to the EVX-B3 deal as such, the EVX-B2 evaluation period has been extended. So 2025 is shaping up to be a pivotal year for Evaxion.

We've achieved several key milestones since the last business update. As mentioned, MSD exercised the option on EVX-B3. We presented two-year clinical outcome data from our EVX-01 Phase II study, and we have announced the addition of EVX-04 to our pipeline and EVX-04 is the lead candidate for our precision cancer vaccine. And looking ahead, we are expecting to provide further R&D and business development updates.

So let's shift focus to our recent R&D and AI-Immunology progress. EVX-04 is our novel AI designed cancer vaccine candidate targeting nonconventional antigens from the dark genome, so-called endogenous retrovirus or ERVs. And these ERVs are specifically expressed in cancers of dormant in normal tissue, making them an attractive target for cancer vaccines. EVX-04 is a therapeutic cancer vaccine aiming to induce immune control in acute myeloid leukemia, where we know relapses remain a major challenge. The vaccine is based on our AI-Immunology platform, leveraging our discovery engine to identify multiple and optimal tumor-specific epitopes that matches the expression profile and also the immune characteristics in patients. And we have now designed the lead candidate and have conducted preclinical studies. Next steps include GMP manufacturing and additional IND-enabling studies to prepare for a first-in-human study.

With the EVX-01 lead vaccine candidate selected, the program has been added to our pipeline. Further changes include the removal of the EVX-02 program as we do not have any active development currently ongoing on this program. The EVX-02 vaccine program serves as proof-of-concept for DNA delivery of neoantigen and has informed on the design of both EVX-03 and 04.

Our lead program, EVX-01, a personalized cancer vaccine that includes multiple patient-specific targets, so-called neoantigens that we identify with our AI-Immunology platform from patient tumor material. EVX-01 is administered in combination with pembrolizumab, an immune checkpoint inhibitor to enhance the clinical efficacy. So, in October, we presented two-year clinical outcome data from our Phase II trial in an oral session at the ESMO Congress. And the results are highly encouraging and were received well by the scientific and medical community. So, at the congress, we reported a 75% objective overall response. We also saw that 11 out of 12 patients that responded had a sustained response at two years mark. We also saw a 34% conversion rate, meaning that patients with stable disease or a partial response deepened their response upon EVX-01 treatment.

So we find the clinical outcome data very encouraging. And further, we believe that they compare favorably to historical pembro monotherapy data. Equally encouraging is the strong immunological activity of EVX-01, which is critical for long-term efficacy. So, in all patients treated with EVX-01, we saw a neoantigen-specific T cell response. Further, when assessing the individual neoantigen immune responses, we demonstrated that 81% of the vaccine neoantigens administered across patients were immunogenic. So this high hit rate provides strong evidence of the predictive power of our AI-Immunology platform.

We also showed that the immune responses were sustained throughout the two-year trial period. Even after dosing after the dosing period ended, T cell activity remained high, indicating lasting immune memory. And durable T cell responses are essential for preventing relapses and in achieving long-term control of melanoma. So this reinforces the potential of EVX-01 as a personalized immunotherapy that not only drives tumor shrinkage in combination with standard of care, but also builds on a robust immune defense.

As mentioned, our AI platform has been enhanced with a new automated vaccine design module, significantly reducing design time and also accelerating development time lines. It enables us to optimize vaccine candidates with high precision, both for new and approved vaccines. So, from data input to candidate generation, the process is now fully automated, ensuring optimal sequence and confirmation of vaccine targets. And as mentioned, the new module speeds up vaccine development while reducing costs compared to traditional methods. And further, it seamlessly connects with downstream processes supporting a smooth transition from design to production.

And the design module has already been applied in some of our key R&D projects. One first example is the use of the module to identify and select regions of an antigen that can be expressed. So we noticed that the full length of a particular antigen could not be expressed due to solubility constraints. But when we applied the new module, we identified truncated variants of the antigens that then could be expressed, opening for preclinical evaluation of that antigen.

Another example is of the application is that we can, with the module, predict most optimal sequences of a vaccine target based on a given antigen protein structure. And here, we have also been able to rescue a HER2 express protein that would require labor-intensive and trial and error design approaches to find expressible constructs. So with this new design module, it position us at the forefront of AI-driven vaccine innovation and further enable us to move fast from target discovery to final product candidate.

So, in summary, we have seen significant progress across our R&D pipeline and AI platform, and we are on track for the next milestones. And we look forward to updating you as our programs continue to advance.

So, with that, I would like to give the word to Thomas to present the financial results.

Yes. Thank you, Birgitte. I am happy to present the financial results for the quarter. And maybe let me just start with an overview of the achievements that we have done throughout the year up until now based on strong execution of our financial strategy.

As we can see here on the slide also, throughout the year, we have made a number of activities with capital market activities with, of course, also the agreement that we did with the European Investment Bank of debt conversion plus also MSD out-licensing of the EVX-B3. So throughout the year and up until end of October, we have activities in the -- to the sound of USD 31.8 million, all which basically helps strengthening our equity and certainly also our runway, which is now extended into the second half of 2027. So really a good and strong achievement and following the financial strategy that we have laid out.

If we zoom in on the third quarter and the highlights from the third quarter, we certainly have had a strong financial quarterly performance. As mentioned, the cash runway now has been extended into the second half of 2027. And we've also seen in the quarter the option exercised by Merck or MSD, which not only provides cash income now, but also has a future revenue income of potentially up to $592 million.

We are well on track also in the third quarter on delivering on our financial targets for the full year. And throughout the quarter, we have also really solidified our equity through the European Investment Bank debt conversion and also through the MSD income. So really a good and strong quarter.

If we look a little bit closer on the profit and loss element of it, clearly, the revenue from MSD drives the first -- drives the quarterly operational gain, the first of its kind for action -- but also importantly, our operating expenses, not only do we manage those well, but we actually also are slightly below last year and more or less at the same level as previous quarter. So also from earlier communications, we expect from a cash flow perspective to still hit around about the $14 million operating cash flow level. Net financials in the quarter is to the sound of $1.3 million, driven by the debt conversion that we did in July. The debt conversion in July with the EIB happened at an 89% share price premium at the market close of July 10. That premium has been recorded on the financial income for the quarter. And that then brings the income for the quarter to the USD 4.6 million.

Turning to the balance sheet. We certainly, as mentioned already, have continued the strong execution that also has improved our equity. The equity now stands at the end of the quarter at USD 16.6 million. Included in the equity is a derivative liability with a net impact of $1.5 million. The derivative stems from our public offering in January and the investor warrants that we had from that date. However, in October, we have seen, as mentioned already, warrant exercises of $2.7 million, which means that this net impact of the derivative will be at a minimal value at the end of the year. So also a good outcome. It's also reduced our outstanding warrants by 1 million ADSs, and we now have a remaining outstanding warrants of 2.8 million. Also important to note is that our cash balance end of June is at a sound and solid USD 10.6 million and we'll see further cash income as we -- or cash flow in as we received in October, the revenue income from MSD and also the cash from the sales of shares due to the investor warrant exercise.

And last but not least, the debt conversion -- debt to equity conversion with the European Investment Bank really has strengthened our balance sheet, has improved our cash flow as we move forward and has certainly also lowered our leverage. So really a great outcome from that event also. So a good solid quarter following along our financial strategy.

And with that, I then hand it back to Birgitte.

Thank you, Thomas. So lastly, as conclusive remarks, I would like to highlight that we do have a strong operational momentum. We have achieved the majority of our 2025 milestones and are tracking towards several potential value catalysts. Business development remains a key priority and multiple parallel partnership discussions are currently ongoing. Cash runway is extended to second half of 2027.

And with that, I would like to thank you for your time and attention, and we'll be happy to answer any questions. Please follow the operator's instructions. Thank you.

[Operator Instructions] First question comes from Soumit Roy at JonesTrading.

Congratulations on all the progress this quarter. A quick question on the EVX-01. What are -- if you can give us any color on the potential partnership deal, like what seems -- what is the key question that you're getting from a partner, they want to wait for a much longer-term data or any other key achievements you have to present for a successful deal?

Yes. Thank you for that question. So we just presented, as mentioned, the two-year clinical outcome data. And I would say that we have moved from questions around quality of data, how does your platform work to more how can we apply your technology within perhaps other types of cancer indications. So the data was received well, and we have been discussing it with key opinion leaders. We have also been discussing it with potential partners. And the strategy is that we will out-license it at the current development stage and then the partner can, of course, decide on potential next step.

If we continue on the track we are currently with advanced melanoma, the next most likely step would be to conduct a larger randomized control arm study comparing the combination of EVX-01 plus standard of care to standard of care alone. So that's at least one option, but we also do see the potential of taking our technology and applying it in other disease or cancer indications where there is a high mutational burden, meaning that we can select high-quality neoantigens and formulate a vaccine that would benefit the patients.

So multiple different discussions are ongoing and also different questions are coming our way. But generally, the questions are not about the quality of the data or the impact of the data, but more how can we move this forward together and find solutions for manufacturing and also for other indications.

Got it. One last question. Congrats on unveiling the EVX-04 in the AML program. If you can give us a slight understanding on the target or how is it -- is it expressed on AML stem cells, like CD33, CD123 have been tried and trying to understand what is the differentiation from this target.

Yes. No, this is a very good question. So the way that we have applied our AI-Immunology platform is that we look at genomic and transcriptomic data. And for this particular type of antigens, we mainly look at transcriptomic data. So the sequences that are being expressed as messenger RNA or RNA in general in the tumors. And then we have looked -- we started out by actually analyzing several different types of novel classes of antigens. And we realized that in certain indications and AML is one of them, there's a very high expression level of these endogenous retroviral sequences from the dark genome, meaning that we can across patients, find shared sequences and put them into a vaccine and thereby being able to support several patients with one single vaccine.

So this is an off-the-shelf approach where we will be able to use the same vaccine across the different tumor profiles and also across the different immune characteristics of the patients. But we're still using AI-Immunology and our core technology to identify the most optimal antigens. Now it's just coming from the dark genome.

The next question comes from Nelson Cox at Lake Street Capital.

Nelson on for Thomas. Congrats on all the progress here this quarter. I'll maybe ask my two upfront, and apologies if I've kind of missed some of this, I had some technical issues. But a lot of updates here as of late. Maybe at a high level, can you please comment on just the overall breadth of partnering conversations you're having across your pipeline and how those have kind of evolved over the last year?

And then when you look at the proportion of your business development conversations you're having today, can you kind of talk about how many are focused on target discovery versus the programs kind of already in your pipeline?

Thomas, thank you for that question. So we do have multiple dialogues ongoing. And I would say that they are across -- the interest is across our R&D pipeline, but also centered around our capabilities for identifying novel targets, so classical target discovery programs. There's interest in our oncology programs, and there's also interest in our infectious disease programs. So it's a little bit mixed, I would say, and some companies do have preferences in infectious disease. Some do have interest in both vaccine candidates that we have developed and in target discovery collaborations. So a little bit of mix and that confirms, I would say, our strategy to monetize on both our own in-house developed vaccine candidates, but also to enter into or target discovery collaborations.

I cannot comment so much on exactly where we are as it's really difficult to speculate exactly on the timing of when these different dialogues would move into a real deal. But a lot of activities, and we can see that the interest is increasing when we have major data readouts as we have had in this last quarter with the EVX-01 Phase II data coming out.

[Operator Instructions] The next question comes from Swayampakula Ramakanth at H.C. Wainwright.

This is RK from H.C. Wainwright. So, certainly, there are very interesting developments going on at the company. So can we focus for a second on your automated design module, which is yet another interesting AI design -- drug design module that you have. So how should we think about this? Is this something that can help your internal designing -- I mean, designing of your internal molecules only? Or is this up for entering into partnerships? Or can you utilize this as a separate licensing situation where any of your partners could take that into their own computing systems and run on their own proprietary molecules. It looks like there's a lot of places where this thing could go. And what are your thoughts on that?

Yes, you're absolutely right. It can go in multiple directions. So, in the past, we have used AI immunology to identify novel vaccine targets. And then these targets have then been undergoing manual processing, ensuring that we could also express them and manufacture them. And this process has been labor intensive. So the ambition was to set up an automated process for this. And we have now been able to launch a new module where several different AI tools are being integrating, enabling us to go from target discovery to product candidate selection very fast.

So that can, of course, be applied to our own programs, but we also do see an option of using this capability to support other companies in ensuring that what they select as their key antigens or in general, key targets that these antigens or targets can also be produced in a cost-effective way. So I do see multiple options for monetizing on this new module.

Okay. So, and then coming into the real world and talk about EVX-01 for more -- for another minute. At SITC, you're planning to present some additional data from the ongoing trial. What sort of data would come out from there? And how would it strengthen your narrative on EVX-01, not only for yourselves, but also for a potential partner, whether it is just on the drug or on the platform, just as you were talking about with -- when you're answering Soumit Roy?

So at ESMO, we presented the clinical outcome data, and we are still in the process of analyzing patient samples. So -- we have collected samples, blood samples before therapy during vaccination and then also as follow-up samples. And all of these samples from the patients are currently being assessed in our own labs. So, we do, of course, monitoring of the EVX-01 induced T cell responses, but we also do deeper phenotypic analysis. So some of this will go out at SITC. We have a poster presentation, but also at future conferences because we have not analyzed all the many samples that have been collected from the patients.

So more to come, more deep dive into the immune profiles of the cells. collected from the patients. And then we also have the extension phase of the EVX-01 trial where six patients are now receiving EVX-01 as a monotherapy. So more data will come from this subset of patients.

Perfect. And then my last question is on the MSD relationship. And it's great to have the $7.5 million. But how much more -- do you still need to give any additional data for the second molecule? Or is it the Merck still has to complete their due diligence on that -- on the data that you've given them in terms of running confirmatory studies or data analysis for them to decide whether they want to spend the other $2.5 million?

Yes. So, for EVX-B2, MSD is currently evaluating the data that we have provided and further, they are in the process of generating some confirmatory analysis. And that was also why the evaluation term was extended. So we expect that they will come back with an answer in the -- or in the first half of next year. But the process is ongoing, and it's, of course, very exciting, and we would love to out-license EVX-B2 to MSD.

There are no further questions. So I shall hand back to you for final remarks.

Yes. And thank you for joining us today. And please do not hesitate to reach out should you have any additional questions. Thank you.

That concludes today's presentation. Thank you for participating. You may now disconnect. Speakers, please stand by.

Welcome to this webinar sharing the 2-year readout from our Evaxion Phase II study in advanced melanoma.

We are very, very happy with the virtual turnout for this call here. And together with the Professor Adnan Khattak, I look forward to sharing results and discussing data and implications of these data with all of you right from this important Phase II study in high-risk melanoma.

So if we look into the next hour or so, we will be covering an exciting agenda. Initially, I will deliver an introduction to Evaxion and introduction to our AI-Immunology platform and briefly cover the data from the Phase I study in advanced melanoma that really paved the way for the Phase II study that Professor Khattak will present. I'll then hand over to Professor Khattak, who will take us through the Phase II design, who will perform a detailed review of the results and also discuss implications of this data. We'll end with a question-and-answer session. And we really encourage that all of you, if you have any questions, please post them. So there will be 2 different ways of posing questions. One is to simply put it into the chat here to type it directly into the chat. Alternatively, feel free to type in queue or question, and we will then ensure when we get to the Q&A session to unmute you, enabling you to post the question directly to myself or to Professor Khattak.

So on that, I'm happy to initiate this webinar with an introduction to Evaxion AI-Immunology in the Phase I data, as previously explained. My name is Benjamin Wolthers. I'm a physician, and I serve as the Vice President of Clinical Development here at Evaxion. And it's truly a privilege for me and a great pleasure to be able to present for all of you today.

So if we look first at Evaxion, we were founded in 2008 here in Copenhagen, Denmark as a truly AI-first company. We leverage our proprietary AI-Immunology platform for vaccine discovery, for vaccine design and for development. And based on AI-Immunology, we've developed a clinical stage oncology pipeline, out of which EVX-01 that we, of course, are presenting today is by far the most advanced project, but where we had several additional projects in the pipeline. Moreover, we leverage AI-Immunology in our infectious disease pipeline, where we target bacterial, viral diseases with a high unmet medical need.

If we look at AI-Immunology and the AI-Immunology platform, this is really at the core of what we do. It's that pumping heart of Evaxion. And AI-Immunology decodes the human system for therapeutic target discovery. AI-Immunology integrates data from the genome, from the transcriptome and from the proteome really to identify optimal targets for our vaccine. As we'll show later in this presentation, AI-Immunology has proven a very, very high success rate in selection of targets inducing potent vaccine-specific T-cell responses and then enabling these vaccines -- sorry, these T-cells to interact with cancer cells and elicit tumor kill.

So if we look at our pipeline here, EVX-01 is our lead candidate, really demonstrates the performance of our AI-Immunology platform and is really our proof of principle proof-of-concept. EVX-01 is a personalized anticancer vaccine, really designed to match the heterogeneity and the high number of tumor mutations that we observe in melanoma.

When you look here on this slide, this is from a similar publication in 2013 by [indiscernible] Nature, where they listed on the X-axis a list of different cancers and the Y-axis the prevalence of somatic tumor mutations. You see that -- on the left side, at least my left side of the slide here, you see the acute lymphoid leukemia, you see AML as well right, with low numbers of somatic tumor mutations, whereas on the other side of the slide, highlighted here in green, you see melanoma, really harboring a high number of somatic tumor mutations. Also of note, you'll see that on the same side of the slide, right, with a high number of mutations, you see indications such as lung cancer, bladder cancer and also importantly, head and neck cancer. So whereas we have done Phase I and II to show proof of concept with EVX-01 in melanoma, there are certainly also other indications with a high unmet medical need where we could see a therapy, such as EVX-01 making a big impact for patient outcomes.

So now turning to the Phase I trial design. The Phase I trial was conducted out of an expert melanoma site here in Copenhagen, enrolling patients with Stage IIIB, Stage IV melanoma, patients were required to have a performance score of 1 and at least one measurable tumor lesion per RECIST 1.1. Additionally, we did not allow any patients to have active brain metastases when going into the study.

If you look here on the slide on the bottom left corner, we have the dosing regime. And when patients were eligible and initiated the study, we initiated checkpoint inhibitor therapy with a dose according to the label. And then we initiated the process of designing the personalized vaccine. This process is projected in the table on the right-hand side of the slide. And what we do is that we obtain material tumor biopsies and also samples from noncancer cells in the patients. These 2 types of tissue are sequenced DNA and RNA sequence, and then we load the data into AI-Immunology. AI-Immunology then selects up to 10 optimal patient-specific targets for the vaccine. The vaccine is a peptide-based vaccine, and this is manufactured, shipped to sites where it is administered to patients. In the Phase I study, we manufactured with a needle-to-needle time of between 6 to 8 weeks and administered EVX-01 for the first time at week 8. This was then followed by administrations at week 10, 12, 14, 16 and 18.

If we turn to the results and some of the learnings from the Phase I study, we saw that very importantly, EVX-01 was safe, well tolerated with only Grade 1 and 2 adverse events. These adverse events being mainly injection site reactions, low-grade fever and fatigue. We saw very encouragingly that 8 out of 12 patients showed an objective response to the treatment combination, so an ORR of 67%. However, we also saw that out of these 8 patients, 7 patients eventually relapsed. EVX-01 was found to induce an immune response in all patients, and T-cell responses were observed towards 58% of the EVX-01 peptides administered. Moreover, as I mentioned before, we had efficient manufacturing of the vaccine with a turnaround of 6 to 8 weeks.

So 3 key learnings came out of this Phase I trial. First, we selected the high 200-microgram dose as the recommended Phase II dose. Second, we evaluated the tumor relapses, and we found also looking at T-cell -- vaccine-specific T-cells over time that we should include into Phase II EVX-01 booster administration, and Professor Khattak will come more into that, but really to administer these administrations to keep the levels of vaccine-specific T-cells sustained at a high level. And then lastly, we took data here from the Phase I study and additional data and updated AI-Immunology with the ambition of even further improving prediction power of AI-Immunology.

So on that, I would like to hand over to Professor Khattak. Professor Khattak serves as Professor of the Edith Cowan University, and he leads the Phase I Trial Unit at the Clinical Research Hollywood Private Hospital in Perth. Moreover, Professor Khattak has served as a primary investigator in this EVX-01 Phase II trial. And also, he recently, this was 5 days ago in Berlin at the European Society of Medical Oncology, Professor Khattak presented the data you'll be seeing now in an oral session.

So with that, Professor Khattak, I'll hand over to you for -- to take us through the Phase II data.

Thank you, Benjamin. Can you hear me clearly?

Yes.

Okay. Thank you, everybody, for dialing in from various parts of the world. I'm dialing in from Western Australia. It's 10:40 p.m. at the moment. So -- and the audience is mixed. So I will keep the discussions sort of generic as well rather than going too much into oncology jargon.

So may I take control of the talk now, Benjamin?

Yes, please do. Please do, Adnan.

Okay. Can you see my slides?

Indeed.

Okay. Excellent. So melanoma is a fairly progressive skin cancer with a fairly high mortality rate, particularly when patients get diagnosed with metastatic disease or Stage IV melanoma. So from 1975, all the way to 2011, so melanoma was considered to be the graveyard of clinical trials, where about 25 to 30 clinical trials were negative. These included trials with chemotherapy, hormonal treatment, biochemotherapy, until 2011 when the first ipilimumab trial came along and immunotherapy was the first treatment to make a significant difference. So roughly about 10, 15 years ago, so maybe somewhere around about 2005, 2010. So the 1-year survival statistic for patient with stage 4 melanoma only used to be about 20%, 25%. I mean only 1 out of 4 patients will see 12-month mark from diagnosis of the Stage 4 melanoma.

But since then, significant progress has been made. So we got first-generation immunotherapy drugs. We've got second-generation anti-PD-1 targeting drugs, and then we use them in combination as well. So this is a Kaplan-Meier curve of different clinical trials of targeted therapy and immune therapy in Stage IV advanced melanoma. So if you can see, so these are the months since randomization and along the Y-axis, we got a proportion of patients alive. So if you look at the 5-year survival statistic now, so with the combination treatment, the brown line, so roughly half of your patients could be expected to live for 5 years or even longer.

And you can see that the curve starts to plateau around about 3-year mark. If you're doing well at 3 years, now we've got latest data even dragging all the way up to 10 years with survival statistics sort of in a similar range. So that's quite a significant improvement in the last sort of decade. The green line represents targeted therapy, which is the other class of drugs that we use to treat patients with advanced melanoma. Many patients develop secondary resistance. And if you look at the progress through a lot of statistics, they are roughly about 35% to 40%, so inferior compared to immunotherapy.

So this -- obviously, the outlook of patients have improved, but you can see 50% of patients benefit, but what about this group of patients who unfortunately die. So how we can help those patients. And obviously, the treatments that we use, for example, the combination of the first and second-generation checkpoint inhibitors, yes, they're efficacious, but they can be quite toxic as well. And none of these immune checkpoint inhibitor treatments here are personalized. So if I get 10 patients, I offer them the same treatment, but they're not specifically designed for an individual patient's melanoma. So I can't be confident. So it works for some. It doesn't work for others. So they are more generic treatments.

And then -- so there's ongoing need for coming up with better and safer treatment options that are more personalized as well, so that we can increase the efficacy to try to help rescue those 50% that unfortunately die in the first year and also to come up with more tailored patient-specific treatments so that we can treat the right patient with the right drug. So in the Phase II trial of Evaxion saw building up on the -- and further consolidating the Phase I data that Benjamin presented. In this trial, we included patients with Stage 3 unresectable or Stage IV melanoma. The bulk of the patients were Stage IV melanoma patients, and they were treatment naive and a typical ECOG-01. So this means that they are fairly fit patients.

They didn't have any active brain metastasis. The primary endpoint included initially looking at response or in terms of the conversion rate, which is conversion of complete -- sorry, from stable disease to partial response or complete response. So partial response means that you see an objective reduction in the tumor volume by 30% complete response, if you see a 100% reduction in the tumor volume and progressive disease, if you see more than 20% increase in the tumor volume. So these are the different figures that we use. So anything less than a CR and less than in between progressive disease is called stable disease.

So the primary endpoint was to see if we can convert stable diseases into a complete response or a partial response, what we call CR or PR or converting a partial response to a complete response to try to see if we can deepen the responses what have been achieved by immunotherapy with checkpoint inhibitor like pembrolizumab, which is one of our mainline treatments that we use. Secondary endpoints included objective response rate, progression-free survival, overall survival and different adverse events and also the neoantigen-specific T-cell response. Exploratory endpoints included a conversion of progressive disease. This means that patients that are not responding to checkpoint inhibitor, can we pull their disease back into either disease stability or response or a complete response.

So 2 types of treatments were used as per standard pembrolizumab monotherapy, what we call KEYTRUDA, that was started at week 0 and given all the way for a total period of 2 years, which is the current sort of standard set across a number of clinical trials or more contemporary trials that we generally treat maximum for 2 years and a standard dosing schedule of 400-milligram IV every 6 weeks was used. And then at week 12, so Evaxion's EVX-01 personalized peptide vaccine, as Benjamin told you the mechanism of action and how it's manufactured.

So that was started at week 12. Six consecutive doses were given in the priming phase every fortnight, 12, 14 all the way to week 22. And then there was a gap after the priming phase and then leading on to the booster phase and 4 doses were given week 30, 42, 54 and 78. And then patients went into sort of long-term follow-up. More recently, an extension phase has been opened up as well for an additional 2 booster doses because of the preliminary efficacy that we've seen and the benefit that the booster doses are providing to vaccine-specific T-cell responses.

So this was the study population. So it's a Phase II, a small Phase II study, 21 patients, 4 screen failed and then 17 patients were enrolled, but 1 patient developed disease progression. So 16 patients went on to the trial. 2/3 of the patients were male and then all patients were Caucasian and with a good ECOG status or their performance -- or their functional status with a vast majority of patients having Stage 4 melanoma PD-L1 expression and BRAF status was sort of 50% of patients had a PD-L1 score more than 5% and then 50% of the patients also had BRAF mutant disease. So generally, we expect 40% to 50% of the patients with Stage IV melanoma to be BRAF positive.

So in terms of efficacy -- sorry, in terms of safety, so from what we know now from the personalized cancer vaccine or individualized neoantigen therapies, they've got very classic toxicity profile, which is fairly well tolerated and very well managed. And primarily, we're looking at no toxicities from Grade II or above. So as oncologists, we don't want to see -- so this is a CTCAE grading criteria, Grade 1, which is very minor toxicity. But when it starts to hit Grade 3 and 4, it leads to major morbidity and Grade 5 obviously, mortality. So we don't want to see anything showing up in Grade 3, 4 and 5 territory. So if you look at in the middle of the table, you can see the Evaxion monotherapy. So the vast majority of the adverse events were Grade 1, and this included fatigue, injection site reaction, body aches, muscle aches, feeling tired and flu-like symptoms, which we expect with vaccines. So -- and then a couple of Grade 2 adverse events.

And then when they were combined, so again, I think the ones that were attributed to both, so primarily Grade 1s and 2s. And then there was only one Grade 3 and 4 adverse event when a patient develop pancreatitis and secondary diabetic ketoacidosis. But from what we know, that side effect profile is not associated with personalized cancer vaccines. It was driven by an underlying pembrolizumab. So I think very much in line with the data that we have from other personalized cancer vaccines, like the mRNA vaccine. So this is very reassuring.

So at week 12, these are the responses that we saw complete, partial stable disease. So the objective response is usually defined as a combination of complete response and partial response, which was roughly about 50% in patients who were treated with pembrolizumab at week 12, means 3 months into treatment. And then once Evaxion was started and then the best objective response rate later on when you combined 15 and 25 was 75%. So generally, with pembrolizumab monotherapy, we expect response rates roughly around about 40%, 45% mark. So seeing the best objective response rate of 75% is fairly encouraging. So it means that you can potentially convert some responses into deeper responses. But obviously, it's a relatively small subset of patients. So we can't make any bold statements, but it's a very good signal.

So this is the waterfall plot. So it is 0%. Anything lower than this line is generally good. Anything above this line is bad. So the green lines represent patients who had a complete response. And so there were 4 patients with a complete response. This curve is not reaching 100%, but this patient had a substantial reduction in their target volume to less than 5 millimeter. But obviously, the disease doesn't disappear, but obviously, anything less than 5 millimeter in terms of the nodal disease is classified as a complete responder and can be put down at 0 millimeter. The blue lines represent partial responders. The gray line represents a stable disease, and there are 3 patients with progressive disease. And this patient did have shrinkage of the tumor, but develop new lesions, hence, progressive disease. The responses were measured according to the standard clinical trial evaluation criteria of RECIST 1.1.

I like this graph. So it sort of tells us how the responses transformed. So where they started at week 0 and then week 12 and how it sort of tells us how things changed by the introduction of the vaccines. So you can see these are the vaccine priming doses and the booster doses. And I would like to highlight like these patients who -- this patient who started with a partial response first and then transformed into complete response. Then again, these 2 patients also transforming into complete response later on. So the argument is some people put forward is that, obviously, these patients were already responding and you would expect, I think I'll show a graph later on, that these patients were expected to have good response anyway.

But I think usually, when patients are responding, they will be responding quite early on in the journey. Generally, in the first 3 to 6 months, you will get quite substantial, or they will achieve their peak response. But here, you can see that some of the complete responders are converting around about week 72, so quite delayed. These are the partial responders that continued and then these are the patients with stable disease that develop a partial response. And particularly if you look at this patient 12, stable disease, stable disease. And by the time the vaccine effect kicked in and the booster doses have been given and there was a spike in the T-cell response, you can see they converted to partial response. So that's what we were hoping to achieve that we can deepen the response on top of the checkpoint inhibitor.

And then the overall responses in terms of the conversion rate, so 7 out of 13 patients. So this means that 3 patients transformed from partial response to complete response and 4 patients from stable disease to partial response, leading to conversion rate of 54% and an overall objective response of 75%. The median follow-up was about 2 years. The median progression-free survival and overall survival has not been achieved yet because the vast majority of the responders up to 92%, they continue to respond to treatment at this state.

What was more encouraging also was that identifying a neoantigen. So neoantigen is like the fingerprint of the cancer cells. So all of us have got their own fingerprint, which identifies us. Similarly, the neoantigens are the fingerprint of the cancer cells. So identifying neoantigen is important. So quantitative effect is important, but also qualitative as well. So not every antigen is going to lead to a significant anticancer immune response. And what Evaxion's AI-Immunology platform identified was that the neoantigen induced T-cell responses in 81% of the cases, which is very encouraging and further built up on the Phase I data where it was roughly about 65% from memory.

This is a very nice graphic representation of how the vaccine-specific T-cell responses happened. So these are both combination of CD8 and CD4 T-cell responses. So this is where the vaccine -- the priming doses started and then the booster doses came along. So if you look at -- this is the pre-vaccination period when pembro was given. This is with the priming phase. You can see the T-cell responses happened quite early on in the priming phase. This means that you start to see an effect because particularly when you're treating a patient, you want to have a therapeutic effect that starts early on because many patients with melanoma were present with high-volume metastatic disease. So you want to generate an anticancer immune response that translates into reduction in the tumor volume so that you can make patients' quality of life better by reducing the tumor burden. So we want a drug that kicks in early.

And then what we also saw more interestingly was the boosting phase, where the 4 booster doses are given, where there was further improvement in these T-cell responses, and they were sustained throughout the period, so highlighting the importance of the booster doses. So -- and that's why I think the further couple of booster doses are being rolled out as well.

So this is a spider plot. So I'll start off with patients who developed disease progression. So the red line. So this patient experienced disease progression, you can see. But if you look at this patient's T-Cell response. So when they were progressing, I think the vaccine had just been started. And then when the vaccine effect kicked in, there was some reduction, but obviously, after that, it sort of plateaued and then further disease progression. And we don't have any further follow-up sort of T-cell data from it.

Then this patient with stable disease, very interesting, one of my patients, you can see the disease volume jumped from 0% to 50%, so almost doubling of their tumor volume at the completion of the first 3-month treatment with pembrolizumab. And then things started to improve when the vaccine was given as well. And you can see how the T-cell responses sort of correlate with that, that when the vaccine has been started, you were seeing increase in tumor volume. And then the vaccine effect kicks in, you see a tumor-specific T-cell response that's going up and the volume of the tumor is going down, down, down, down. And then because this patient had developed disease progression at the beginning, so we can't call them a responder. But we have done a couple of PET scans for this patient now, and they've shown a complete metabolic response, but obviously, as RECIST criteria. So I think we still call them a stable disease.

In terms of the deepening of the response, so I think you can see these patients were responding. But then after that plateaued, so the vaccine effect sort of kicked in roughly around about this period, and that's what we start to see these patients coming down, transforming into complete responders. And then the trajectory, if you can see also here, so these patients in the blue line, these were -- this is a 25% line you can draw here. So at 3 months time point, these patients had stable disease. They hadn't crossed the 30% and then they sort of continuing the stable disease up till here and then they start to respond after sort of the 36 weeks time period.

Obviously, these are a small number of patients. We need to further study the T-cell vaccine data. And obviously, ideally have a bigger cohort of patients to draw more meaningful and more statistically robust conclusions. And this patient, who developed quite a late complete response, you can see how the tumor volume started to go down when the T-cell response went up and it really spiked, you can see the tumor kinetics hit 0 transforming into complete responder.

So in conclusion, I think before we look at the efficacy data, we always want to make sure that whatever treatment we roll out that is safe to use. So this combination was well tolerated with a good safety profile, consistent with the Phase I data, objective response rate is very encouraging. But again, it's a small subset of patients. But what is quite encouraging is that most of the patients, 92% of the responders demonstrated sustained response at 24 months time point. And then more interestingly, the ability of this combination to induce potent and sustained T cell responses. And then the AI platform's ability to recognize the new antigens that will generate anticancer immune responses in 8 out of 10 patients. And the vaccine was successfully manufactured for all of these patients.

So we don't want to screen fail. I mean, I'm involved with other personalized cancer vaccines like mRNA vaccine. So we do have a certain vaccine failure rate, particularly the recent clinical trial that we concluded. So this is very encouraging that in this small subset of patients, nobody failed screening because of an inability to make the vaccine. So these findings are very encouraging. They give a very positive signal, and they support the ongoing development of Evaxion in high-risk melanoma and further discussions are ongoing by Evaxion's team. So thank you.

Thanks a lot, Dr. Khattak, for this comprehensive overview of the data and running through this. I think this takes us to the Q&A part of the webinar here, and I'm happy to see that we have quite a number of questions here. I think I will start by -- I have so far looked at 3 people who have their hand raised. And maybe we should start with Scott trying to -- we will ensure that you should be able now, Scott to unmute and post your question to Dr. Khattak or myself.

You're still muted, I believe, Scott. Maybe -- while we figure this out, I'll go to one of the questions in the chat. This is from Wim, who asked whether EVX-01 -- if EVX-01 was licensed to another company, how dependent would that partner be on the vaccines infrastructure, considering its personalized therapy? How do you see this working in practice?

So thanks a lot for this question. This is, of course, something that we're discussing with relevant partners. And without disclosing any of that, I will say that if we break down the designing of the EVX-01 vaccine, the sequencing part is very straightforward in many locations, such can happen decentralized at the site. The predictions AI-Immunology is, of course, the core, as I mentioned, of what we do. This is after being developed, we have a 24-hour turnaround time on this, making it quite really removing any complexity from this.

And then I'll also say that the manufacturing of the peptide-based vaccine is using very conventional peptide manufacturing production methods. So in that sense, not complex. However, how this will work when we move forward into our development of EVX-01 with a partner that is really, of course, dependent on discussions between Evaxion and that partner.

I hope that suffice.

I don't know whether Scott has been able to unmute. If not, then let's try to go to -- I have a RK listed here as having his or her hand up. So maybe RK, if you are able to now unmute and pose your question. Let's give it a shot.

Dr. Khattak, first of all, thank you very much for doing this call and talking to us this late in the night. So a couple of quick questions.

The first one being the ORR, which you report 75% in the 16-patient study. Certainly, it looks better than the 40% to 45% that we have seen with the monotherapy. But however, how would you think about this response rate based on not just what has been approved, but other therapies that you have worked with? And what does this say or what should we be expecting as we go into the late-stage studies with the same molecule?

Yes. So I think very, very good question. So I think our current first-line treatment options across Australia and across the globe, obviously, this is a very busy area. A lot of new clinical trials are happening. Different molecules are being added to each other. But unfortunately, they also bring a lot of toxicity as well. So one of the most common regimes that we use is a combination of ipilimumab and nivolumab. So where we expect response rate somewhere around about 50% to 60% with the combination, but it's fairly expensive treatment, number one. Number two, it can lead to substantial toxicity where you got to -- you probably saw that when I mentioned that we do not want to see a high Grade 3 or higher adverse event rate.

Now that Grade 3 adverse event rate with this combination hits about 50% to 60%. This means that you expect that 50% to 60% of the patients are going to end up in the hospital in one way or another due to some sort of toxicity. So we got a very good efficacious regime, but it's fairly toxic. And obviously, quite a substantial of our patients are 65-year plus. So we then we tend to shy away from using treatments like this. So certainly, we want to cross that 60%, 70% mark and try to approach more into that territory, 75% plus, but at the same time, not lose the -- in terms of safety.

Now tumor-infiltrating lymphocytes are being rolled out as well in the first-line setting. I'm involved with those trials as well. But that's quite a cumbersome process. And at this stage, in the melanoma community field, we're not sort of entirely convinced that the cell therapy has got any striking role in the first-line setting when we got other good efficacious options available that are off the shelf, and we don't have to go through a complex process of cell therapy manufacture.

The personalized cancer vaccine field is very interesting that, obviously, we're seeing improvement. So the personalized mRNA vaccine data that presented at ASCO a couple of years ago, that was very encouraging in the Stage II early-stage melanoma setting. And now we're waiting the results for the Phase III from the early-stage melanoma setting. But doing them in the metastatic setting pose a challenge because we need to obviously start treatment, but at the same time, because these are not off-the-shelf treatments, they need to be manufactured.

And then we don't want to be in a situation where you start a patient on the standard treatment plus send up the tissue to your central lab to do all the DNA, RNA sequencing and come up with a personalized vaccine and then find out 6, 8 weeks later that we failed to manufacture the vaccine because I've seen it in other clinical trials, vaccine failure, and I've had recent experience in the context of another trial where we waited for a good 6, 8 weeks for a patient to be even randomized because you can't randomize patients till you obviously have the approval or the indication from your lab that the vaccine can be manufactured.

And then unfortunately, for those couple of patients, we could not obviously proceed. So you don't want to waste that time. And then at the same time, try to deliver the vaccine as early as possible so that you can start your priming phase early so you can generate your T-cell response. And then at the same time, the data from Evaxion is quite encouraging with the 100% manufacture success rate. Obviously, it's a small population need to be replicated in a bigger, ideally a randomized Phase II study, but this is very encouraging that the toxicity profile goes in its favor, the manufacture and I think the translational response in terms of patient-specific and durable responses are very encouraging.

Can I ask a follow-up?

Yes, maybe a quick follow-up question.

Okay. The follow-up is the 4 patients that did not respond, was there anything unique about these 4 patients either in terms of failure of the vaccine itself or in terms of meaningful T-cell response in these individuals?

Yes. So I think, obviously, the different -- obviously, it becomes hard to drill down to an individual patient's data because they don't have access to each individual patient's data as an investigator. But eventually, obviously, we're looking into those details. But yes, I think I saw the patient where they had disease progression. So we couldn't have the T-cell responses sustained in some of them.

I'm not sure if Benjamin, you want to comment on that as well. So the others that progressed, what was their individual T-cell response because I don't have all those curves. Maybe I think the backup slides, you think you have something on that maybe?

Yes. So indeed, I can say that we are, of course, drilling down, right? The data presented today is with the data cutoff of end June, right? And we've been cleaning data, getting everything ready for ESMO in this presentation. But over the coming weeks to months, we will, for sure, be evaluating together with Professor Khattak and also our other key opinion leaders that have served as PIs in these studies, whether we can decide for any trajectories in responders and nonresponders. But to Dr. Khattak's point, right, we had 12 out of 16 being responders, right? So they are small numbers, and we see T-cell responses in all patients, all 16 patients.

I will go to the next person I have here on my list that is Thomas Flaten. So Thomas, if you are able to unmute, please do so and post your question to Professor Khattak and myself.

We may just wait until we get the connection there.

Then I have Kevin on my list as well here, who has raised his hand wanting to ask a question. If there's any of you who have a question, but are not able to unmute or whatever for technical reason, feel free, of course, to type in the question in the chat. We will take those after this question from Kevin.

Kevin, are you able to unmute and please post your question. Maybe not. If that's the case, please type in your questions in the chat. We will take them in this way.

Benjamin, do you think maybe everybody has been muted by the organizer. Is that the reason because I think Thomas pointed that out.

But we are able hear in the background...

Can you hear me now? Sorry...

Yes.

I think I broke the code. Yes. So listen, it's a wonderful discussion. I just want to -- maybe one follow-up and then a question. I appreciate your responses to RK in the context of what else is out there in terms of other potential new therapies. At ESMO, the primary focus in other tumor types was really sort of the ADC space. Just do you have any comments with regard to ADCs and melanoma? Is there a path in your view? Or can you maybe just add that and then I maybe have a more substantive follow-up.

Yes. So I think, yes, ADCs are very topical. Obviously, there are different bispecific antibodies. There are T-cell engagers that are being used in melanoma. But again, obviously, these are not necessarily patient-specific. There's no proven role of ADCs at this stage, but a number of clinical trials, Phase I and IIs are looking at melanoma subset, particularly in PD-1 refractory patients. Now some of them are purely sort of chemotherapy or put it that way a fancy version of chemotherapy drugs type ADCs, where you use a [indiscernible] inhibitor.

But then what we're seeing more promising now is the anti-PD-1 and anti-VEGF bispecific antibody like, for example, ivonescimab, which had some really good results in the lung territory recently. So those type of molecules certainly hold also promising feature in this field moving forward. But again, I think this will be in not a patient-specific, patient tumor-specific territory where we would expect that some patient will respond and then a certain patient population would not respond. So it will be wiser than what we already achieved with other checkpoint inhibitors so far.

In terms of clinical trials, I think I've just become aware at this ESMO meeting that some sponsors are looking into this territory in the early phase melanoma setting in the perioperative setting, for example, patient -- not everybody responds in the neoadjuvant setting to the checkpoint inhibitors to see can we rescue the patient who are nonresponders after the neadjuvant treatment? Can you add in an antibody drug conjugate? So this area is going to be explored and quite interesting to monitor moving forward.

And then my sort of question specific to the Evaxion data, which is very interesting. It's just -- do you think we've learned anything in this most recent data set with regard to kind of dose and schedule. I mean, on adverse events, clearly, it looks like if there was a rationale for going with a more intensive regimen either on dose or alternative dosing frequency, it seems to have that flexibility. But when you kind of look at the patient responses and specifically maybe some of the T-cell data in the patients that you've seen, is there -- are there kind of improvements that might be captured through schedule change and improvement?

So do you mean in terms of changing the dosing schedule or starting vaccine early potentially or...

I mean in terms of, for example, a more sort of intensive boosting window or maybe having the option to go to more frequent dosing in the first year and then to a kind of a different sort of taper in year 2 and beyond. I mean, I think that's...

Yes. I think a number of possibilities can be explored. So I think probably somebody posted a written question that, could we start the vaccine earlier? So yes, I think we waited for 3 months. This doesn't mean that we didn't have the vaccine ready for 3 months. So -- but it makes a smaller study like this very cleaner, so that you can tease out difference between what vaccine is contributing compared to what pembrolizumab has done. So if you were to do everything together and somebody gets a response, it becomes hard to tease out how much the PD-1 contributed, how much the vaccine contributed.

So ideally, you obviously have a Phase II of the vaccine combined with pembro versus pembro alone. So to tease out this question, what is the vaccine adding? So like what we did in the KEYNOTE-942 trial, it was in the early-stage melanoma setting where a personalized vaccine was given in combination with pembro versus pembro alone to sort of answer a similar question.

In terms of the frequency, could we change that? Certainly, yes, we can. But obviously, every dose that you give has got a dollar amount associated with it, unfortunately. So how much that would contribute? I mean, we've got a good schedule. So if you look at, for example, Moderna's mRNA vaccine, all the clinical trials, the INTerpath trial that they're doing in melanoma, RCC, lung trials, perioperative trials, they're giving a total of 9 doses. There's no booster doses involved.

So Evaxion is the first trial that I'm involved with, where a boosting dosing schedule was rolled out and has gone through an efficacy and then further extension boosting has been rolled out, very much in alignment to your feedback because we didn't plan the additional booster doses until we stumble on this that when we saw that we're seeing a therapeutic effect and we're seeing a vaccine-specific anticancer T-cell response as well in others. So we're learning as we're going along. And yes, things can be certainly modified and tweaked and this is how we will learn.

So -- but obviously, at the same time, I think from a financial perspective as well, I think we just have to make sure we get a good rationale for adding a particular treatment. For example, we saw that there was boosting benefits seen and the boosting effect seen hence further additional 2 booster doses were added. So we'll probably get an idea in terms of the booster doses. Do we need to spread it out more? Or do we need to do it more frequently to begin with? So these things can be certainly explored down the track here.

So I see Thomas right, and I apologize for you not being able to speak here, your question, but I'll try to read it through. I think the first one, Dr. Khattak, it's for you, right. Can you comment on any potential read-through from this EVX-01 data into previously treated melanoma patients, particularly in light of the enthusiasm for the Moderna program?

So is this question on the chat?

Yes, this is in the chat.

Any idea speculation to begin with vaccine earlier. So this a question number from.

This is from Thomas Flaten, where it says, although speculative, can you comment on any potential read-through from this study into previously treated melanoma patients? So any read-through from this study into relapsed/refractory...

I think. One, obviously -- so one option is obviously that there will be a group of patients who would respond beautifully to pembrolizumab monotherapy. So not everybody needs doublet treatment. It's just that our understanding that biomarker is not at a level that we're able to tease out and identify. So as a clinician, what I find very frustrating when I was speaking to the patient is that we make it fast, we make a big deal that we got these drugs that people are living a 50% survival rate at 10 years for Stage IV melanoma amazing. But I can't tell that individual patient, are you going to be in the 10% to 50%, who are going to live for 10 years or not. So certainly, our understanding hasn't improved to that level. But I think rescuing patients who have not responded is certainly a good area of exploring further benefit from these vaccines.

The other option that I was also thinking was that in the, for example, neoadjuvant treatment or perioperative treatment is very topical and has changed practice. And that's with the rollout of adjuvant and neoadjuvant treatments, we're seeing much less Stage 4 and routine clinics now. And one area particularly that of interest to me and other clinician is that patients who do not respond to preoperative checkpoint inhibitor treatment, what do we do with them, the nonresponders? Do we escalate their care to combination ipilimumab and nivolumab? Do we come up or if they already received the NADINA trial protocol of ipilimumab and nivolumab and they don't respond, where do we go from there? What do we do with those patients? So that's another area where the personalized vaccine could be explored because we'll have access to good quality tissue in the postoperative setting. So yes, that can be certainly explored as well.

Thanks, Dr. Khattak. And Thomas, you have one further question addressed to me. It reads are the boosters the same as the priming doses? And has any thought been given to evolving the boosters to adjust for immunogenicity of the new antigens chosen for the priming doses?

And maybe taking the first question first. It is the same dose, right? So when we manufacture EVX-01 based on, as mentioned or described in detail earlier, right, that really gives us one patient-specific vaccine. It's the same vaccine targeting the same neoantigens that we use throughout therapy, and we also use the same dosage throughout therapy.

And then has any thought been given to involving the boosters? So what we see and it's not shown here, right?

As I said, we're digging into the data and we'll elucidate more is that we see over time, we see more and more of the antigens, the neoantigens in the vaccine eliciting vaccine-specific T-cell responses. So in that sense, I don't know if you're alluding to redoing a biopsy to test whether the neoantigens are the same. However, that is not currently with the response rates and also the duration of responses presented by Dr. Khattak. That is not something we for now at least see is needed to provide an impact on trajectories for these patients. Thanks a lot for the question.

I think then I don't know -- I have Kevin still maybe with his hand up here. Kevin, I don't know if you're able to unmute on this, then I'll go up score a little bit up here and see there's a question from Thomas. Thomas Malten saying, how does EVX-01 compare to the IO Biotech plus nivolumab therapy that was also shared here at the recent ESMO? Are regulatory challenges more pronounced for personalized vaccine relative to off-the-shelf vaccine? Are there differences in duration of outcome? Is there room for optimizing EVX-01 during development? And was there any signs of this during the Phase II study?

So I'll try to chip away at them, Thomas, right, and then comment if you're missing something, right? So compared to the cancer vaccine developed by IO Biotech, we're very, very different. They have an off-the-shelf vaccine IO-based really targeting regulatory T-cell compartment and off-the-shelf in no way personalized to match personal private neoantigens in these patients. So in that sense, or being a vaccine technology, the concept is very different between IO Biotech's approach and the one we are using at Evaxion.

I'll say regarding regulatory challenges, these regulatory interactions is, of course, something we're having on an ongoing basis. When we initiate the Phase II study here, we submitted and discussed with regulators across Australia, Europe and U.S. And without going into too many details, I can say that regulators were very overall positive in regards to allowing us to conduct these studies. And thus, we have not seen major hurdles there. Also, this is an approach where, as mentioned before, we produce the vaccine, we're able to test the vaccines as we should before administering this to patients. Maybe this is so very much in differentiating this from some other cell therapies needing fast infusion.

Are there any duration in outcomes? As highlighted here by Professor Khattak, we see 12 patients responding to therapy. One of these patients died from an unrelated intracranial hemorrhage 1.5 years into therapy. Otherwise, all of the responders remain in response. And as Professor Khattak said, this is really encouraging, right, because it's a hint of, that's what we're doing here, right? Do patients live longer? Do they live better? We're seeing these sides of them living longer here with this therapy.

So in that sense, we see no differentiation in these patients and room for optimizing EVX-01. I think Dr. Khattak went into that before in regards to dosing, et cetera. I will, however, say that from discussions here at ESMO with key opinion leaders, et cetera, the results are promising and the effect of the boosting vaccinations compared to what we see in the Phase I without boosters is really a tremendous, right, the level of sustained T-cell responses there.

Good. I hope that suffices.

There's a question from Juhi asking, discussions are underway with regulators. On the last page, means that there will be a possibility to make contracts with your partner companies. And of course, we are in discussions with both relevant partners who find this data interesting and wants to move this forward into further clinical testing. And in regards to the regulatory discussions, as a clinical developer, you know that it's always best to be in close contact to regulators across the world, and we are doing the same by planning for further interactions to discuss this data and possible next steps with the relevant regulatory authorities.

Then maybe there's a question for you and Dr. Khattak. Were there any differences in responses based on BRAF or PD-L status?

I don't think. So we evaluated the data specifically based on BRAF status because obviously, 50% of the patients have BRAF mutant disease. So you're looking at 7, 8 patients, so we can't draw any meaningful conclusions from that. And obviously, we don't necessarily use PD-L1 status as a routine biomarker in managing patients with advanced melanoma.

And then I think the next question is that based on Slide 20, almost all responses were observed before initiating EVX? So I think it's more looking at the trajectory where they started. Yes, I think there were some responses that were happening, and that's obviously the response rate was 50%. But I think it's more in terms of the deepening of response with time. So those who hit the partial response in the first 12 weeks, if you were to follow their tumor kinetics for a bit longer time. So they were not all going to transform into complete response. So you can see some of the spider plots, this initial response and then the curve is flattened out for some period of time and then before it converts into a complete response.

And similarly, those who started as a partial responder and then later on -- so those who started off with stable disease at 2-year time point -- sorry, 12 weeks time point, they transform later on. And I think one of that -- the little BD diagram with the different beads, which I showed. So usually, you would expect in clinical practice that we see patients sometimes have a delayed response, but they usually achieve their peak response within the first sort of 3 to 4 months in the vast majority of cases.

And as I showed, I think in that curve, one of the patients who had stable disease and transformed into a partial responders quite like after week 72. And then a couple of patients who were partial responders to begin with, if you look, follow those dots in those slides, so they remain as partial responders up to week 60 and then another one up to week 72. So that's like 1.5 years into your treatment. We generally don't expect partial responders certainly becoming complete responders so far down the track.

But again, yes, certainly, this can be by chance. And hence, the need always we look at statistical significance and to put the matter to breast, how much the vaccine is adding. Yes, you do a randomized trial to see how much is contributing. Then you will have your PFS, progression-free survival data, your survival data, your response rate. So you get all those sort of things figured out. But obviously, in a smaller population, we're just looking at more signal generation rather than robust evidence as such.

Thanks a lot. I think we're almost at time. However, there's one -- there's a question from Jan Stenvang here that asks, any ideas or speculation to begin with the vaccine earlier? Any idea to treat only vaccine and not pembro?

And I'll say that indeed, right? I think Professor Khattak, you did a good job explaining why we started at week 12, right. For Phase I, we initiated therapy at week 8, right? But indeed, this can also be moved earlier also depending on the trial design and further evaluations of the efficacy EVX-01. And last comment maybe to treat with pembro only, I think this also comes back to one of the aspects raised by Dr. Khattak from earlier, right, that, of course, the combination therapy will also depend on prior lines, which line of therapy you go into, what makes sense for what patients. And those discussions is, of course, discussions we are having with -- we'll be having with Professor Khattak, with partners and of key importance also with regulatory authorities.

And then as maybe I think we are at time, and I see maybe there's a few additional questions from Andrew. What's the question? Thank you. So I see a few additional questions. I will reach out to you personally to answer these questions in text with 2 minutes over time.

And on that, I will say thanks a lot all of you for attending this webinar, in particular, thanks to you, Adnan Khattak, for staying up late, right, and trying to speak with not too much loudness so that you wake up your girls, right, and we'll be in touch, of course. All the webinar slides here and the recording will be available on the website, right? If you have any further questions, please reach out to the Investors team.

On that, have a great evening, day or morning wherever you are, right, and take care. Bye all of you.

Pleasure is mine. Bye-bye.