Good afternoon, and welcome to the Elicio Therapeutics Virtual KOL Event. At this time, all attendees are in a listen-only mode. A question-and-answer session will follow the formal presentations. As a reminder, this call is being recorded, and a replay will be made available on the Elicio website following the conclusion of the event. I'd now like to turn the call over to Bob Connelly, Chief Executive Officer at Elicio Therapeutics. Please go ahead, Bob.
Thanks, Tara, and welcome all to our virtual Key Opinion Leader event, where we will be discussing our amplifier-powered ELI-002 cancer vaccine for the treatment of mutant KRAS-driven pancreatic cancer. In your next two slides, two and three, we're going to download these slides, the entire deck, to our website as soon as the presentation is over, and I would encourage you to look at our standard safe harbor and forward-looking statement disclaimers. We can go on to slide four. I'll do a quick intro to the agenda today and then our guest speakers, and then hand it off to the guest speakers, and then I'll wrap up at the end of the session today. We will initially kick off with Dr.
Dr. Darrell Irvine, who will talk about the challenges that cancer vaccines have and are facing, and how lymph node targeting is addressing many of those challenges and enhancing the performance of cancer vaccines. We will hand it off to Dr. Eileen O'Reilly, who will do a relatively deep dive into pancreatic cancer, or PDAC, looking at both the adjuvant and neoadjuvant stages of disease there, what are the current standards, treatment paradigms, and then moving into new advances and what the future looks like for opportunities in KRAS immunotherapies. Chris Haqq will give you a quick summary of what we've seen in phase one of two phase one studies with our ELI-002, talk about the phase two trial protocol to design, the upcoming interim data that we expect in Q3, and a high-level statistical plan for that interim analysis that will take place.
We want to leave plenty of time for questions for the audience, and so we'll take those first, and then if there's time left, I'll do a quick wrap-up here. We can move on to the next slide. I'm joined on the call today by the senior executive team from Elicio, who are all available during the Q&A session, but I want to introduce our first speaker, Darrell Irvine. Darrell is a world-renowned expert in vaccine technologies and specifically in cancer vaccines. He's also currently the professor and Vice Chair at Scripps Research Institute.
After joining Scripps, I believe last year or before that, Darrell had a long tenure at the Koch Institute at MIT and has really dedicated his work and the work of his team to identifying the challenges and the hurdles that vaccines, and specifically cancer vaccines and cancer immunotherapies, continue to face and developing solutions to those problems. One of those solutions was the Amplifier platform, which is the lymph node-targeted approach that Elicio was founded around and which ELI-002 is based on. Darrell will be followed by Dr. Eileen O'Reilly. Dr. O'Reilly is a Winthrop-Rockefeller Endowed Chair in Medical Oncology at MSK. She is a global thought leader and researcher in pancreatic cancer, and we are very honored to have both Darrell and Eileen here to talk to you today. With that, I will hand it over to Darrell.
Thank you, Bob. I'm going to take you through some of the original challenges that we were thinking about in cancer vaccines and some of the hurdles that we identified in early preclinical studies and the solution that we came up with to address some of these challenges and some of the exciting biology that we learned about this technology as we went along. There are several challenges that therapeutic vaccines have faced in cancer. Some of these are well-known problems that have existed since the field started, such as the challenge of selecting suitable antigens expressed by tumor cells that aren't restrained by tolerance and that are broadly expressed by tumor cells. Many early clinical trials of cancer vaccines were trying to treat patients with very heavy burden of disease that undergo many rounds of prior therapies and employed inefficient vaccine regimens or poor vaccine adjuvants.
An issue that we identified in our work early on was the challenge of getting the vaccine to the command center of the immune system, the lymph node. In general, peptide vaccines that are one of the most popular modalities for developing therapeutic cancer vaccines in general have just shown very poor potency in humans. Next. These vaccines based on peptides have many attractive features, however. Generally, these are often developed as 20-30 amino acid long linear epitopes that would contain, for example, a mutation that's expressed by the tumor. They're combined with various adjuvants to stimulate the immune system to respond to that peptide antigen. These have been shown to be very safe. They're relatively inexpensive relative to other sorts of cancer therapeutics. They enable targeting of unique antigens like post-translationally modified peptide antigens.
It's been shown by a number of teams that peptides can be manufactured rapidly, enabling patient-specific vaccines, as first demonstrated by the team led by Cathy Wu and Nir Hacohen at the Dana-Farber Cancer Institute. Peptide vaccines have also faced a number of challenges around finding effective adjuvants for use in humans, issues around the biology of antigen presentation in humans that I'll skip for today. At the end of the day, really, the major limitation has been the limited potency of peptide vaccines in humans, especially in their poor ability to elicit cytotoxic T-cell responses or CD8 T-cell responses that we think are really crucial for addressing cancer. Next slide.
Where we started more than 10 years ago now in my lab at MIT was trying to address what are the key issues that govern the potency of peptide vaccine and what limits peptide vaccine potency. I had recruited a really terrific postdoc by the name of Haifeng Liu, who was a chemist by training. He first attacked this as a physical chemistry problem, and he asked, when we inject a vaccine in the tissue, where does it go? We inject these vaccines typically under the skin or in the muscle, as with all sorts of vaccines that we use in humans, but they need to traffic through lymphatic vessels to draining lymph nodes to reach the key immune cells that respond to that vaccine.
In early experiments, next, in mice, where we fluorescently labeled peptide vaccines and injected them into animals and then took draining lymph nodes out at different time points after vaccination. Here you're seeing an experiment where we're imaging whole excised lymph node tissues shown in the images at the right, and we're trying to detect the vaccine signal. You can see compared to the positive control where we just inject the vaccine directly into the tissue outside of the animal, that there was very little signal being detected in these draining lymph node tissues, saying that a major problem right off the bat is that the vaccine doesn't go where we need it to go. Next slide. As we dug into this problem, we realized this is governed in large part by the physiology of how solids are transported out of tissues.
When you inject a drug into connective tissue, it can either leave the tissue by going into blood vessels, which is not where we want a vaccine to go, or it can go into lymphatic vessels, which is where we need the antigen and adjuvant to be traveling. To first order, that decision of whether the material we've injected goes into the blood or the lymph is governed in large part by molecular size. If our injected compound is small enough to diffuse through the basement membrane and tight junctions that line the blood vessels, then because the blood clears tenfold more fluid per unit time than the lymph does, small compounds will quickly be cleared out in the blood. By contrast, larger macromolecules or particles will instead traffic into lymph vessels, which is where we need the vaccine to go. Next.
If you measure the amount of injected material that goes into lymph vessels as a function of molecular weight, you're injecting a series of globular proteins of increasing size. You see this very clear relationship shown on this graph where small molecular weight compounds have very low uptake into the lymph because of this transport effect. This exactly explains why peptide vaccines are not trafficking to lymph nodes. These compounds typically have molecular weights of a few kilodaltons. They're too small to be trafficked to lymph. They're diluted into the blood and quickly cleared in the kidneys or degraded. By contrast, larger macromolecules, the bigger your cargo, the more likely it is to traffic into lymph. One of the things we noticed is that the most prevalent protein in interstitial fluid, albumin, is just the right size that it traffics one way.
It comes out of the blood, travels through the tissue, and then exits the tissue through lymph vessels. That inspired an idea to solve this vaccine trafficking problem by using the albumin that's already present in your tissue as a chaperone to redirect vaccines to the lymph node. Next slide. That led Haifeng to design what we called amplifier vaccines or AMP vaccines, where we would link either molecular adjuvants that stimulate the innate immune system or peptide antigens to a phospholipid tail that was designed to bind to albumin. Here we're exploiting one of albumin's main functions in biology. Albumin, as shown in this structure in gray at the top, has seven different fatty acid binding pockets. It serves as a lipid transporter as one of its main functions in vivo.
We designed these amplifier peptides so that they would, on injection, associate with albumin through that lipid tail binding to one of these lipid binding pockets. Next. On injection then, these lipid tails would bind albumin and thereby get redirected following albumin's natural path from the tissue into the lymph and thereby redirecting to the lymph node. Next slide. There are a couple of sort of design rules to these molecules that we worked out, and I'll summarize a lot of that biochemistry in just one set of data here. The first design rule is that the lipid tail, we found the more hydrophobic lipid tail with a saturated lipid chain or more hydrocarbons, the more hydrophobic the lipid tail, the more it would bind to albumin in vitro, and that predicted better and better lymph node targeting in vivo.
You see here a series of experiments where we're taking lymph nodes out of mice that have been injected with vaccine that's fluorescently labeled, and we're increasing the number of carbons in the lipid tail, and you see that the more carbons we have in the lipid tail, the more lymph node uptake we see. On the other end of the molecule, because some of these peptide antigens that we might select could themselves be quite hydrophobic, we linked them to the lipid tail through this water-soluble polymer spacer shown in blue, a polyethylene glycol chain. To optimally promote lymph node targeting, we needed a certain number, a certain length of that polyethylene glycol spacer.
You can see this experiment at the top where we're increasing the number of ethylene glycol units in that polymer spacer, and the longer that peg chain, the more lymph node uptake we see. Once we optimize these two design rules, we could really promote dramatic changes in lymph node targeting of either adjuvant compounds or antigens. Next slide. Interestingly, you know the field has pursued lymph node targeting through a variety of strategies, more or less trying to exploit this size-mediated trafficking in the past. For example, liposomes and nanoparticles have been proposed to promote lymph node targeting through having optimal size for lymphatic trafficking.
In a meta-analysis we carried out a few years ago comparing a number of published studies testing different technologies for lymph node targeting, we found the graph you see here, which showed that as you have transport strategies based on smaller and smaller vehicles, with albumin being a very small transporter of about 5 nanometers in size, you see greater and greater lymph node delivery of vaccine to antigen-presenting cells, dendritic cells in the lymph node that are the key cell type to initiate antigen presentation and the immune response. You can see that among all these different technologies that have been tested, the small particles do better than liposomes. Albumin hitchhiking, as we're using with the amplifier, does quite a bit better than small particles. There is this sharp drop-off.
If you get down to small molecule reagents like the free peptide antigen, you lose that lymphatic delivery. We think that this optimal lymphatic delivery comes from the fact that having a really small transporter like albumin means that the AMP vaccine is efficiently traveling through the tissue, making its way through extracellular matrix to reach the lymph vessels in a manner that's more efficient than any slightly larger particle, which will be hindered as it diffuses through the tissue and can get caught up in the tissue before it makes its way into the lymph vessels. Next slide. Now, over the past 10 years or so, we've shown that this strategy is useful for redirecting a whole variety of therapeutic cargoes to lymph nodes, everything from small molecules to peptides to proteins to nucleic acid adjuvants.
There is a generality here that this can be applied in many different ways. As I have already shown you, we are using it at Elicio to deliver both peptide antigens and adjuvant molecules. Next slide. Once you have optimized these design rules that I was talking about, the effects can be pretty dramatic on where these vaccines get to. If we take a vaccine antigen labeled here with a green fluorofluorine injected into mice, where the free peptide shows almost no accumulation in lymph nodes shown here in histology, an amplifier form of that same molecule injected at the same dose shows really impressive accumulation in draining lymph nodes. From these successful experiments in mice, we then turned to ask whether this technology would also work in larger animals that would have anatomy more like humans.
In studies in rhesus macaques, we showed that if we inject fluorescently labeled vaccine and compare a soluble vaccine injected in macaques and look at uptake in draining lymph nodes, here we are imaging whole lymph nodes that have been taken out of the animal after injection. You see on the right that the soluble vaccine shows no peptide signal in the lymph nodes and very little signal from this molecular adjuvant CpG, whereas the AMP vaccine form of both the antigen peptide and the CpG adjuvant are both showing strong delivery to the draining lymph nodes. Further, we carried out positron emission tomography imaging studies in macaques where we could image in the whole animal where these vaccines go. I am showing you one image from a PET study here on the right.
The X marks the spot where the vaccine was injected into the thigh of this animal. You see the dark signal shows where the vaccine is going, illustrating that we not only reach the nearest draining lymph nodes, which are that first blob of signal above the injection site, but also a number of additional lymph nodes that are in a chain traveling up the thorax of the animal. This is one of the other, we think, advantages of this technology. It delivers the vaccine to multiple sets of lymph nodes draining the injection site, and therefore we think giving the immune system a better opportunity to respond to that injected vaccine dose. As we continue to study this technology, we discovered that there are two other additional important features of this particular strategy that we chose to mediate albumin hitchhiking.
The first is that because we linked the antigen to the lipid tail through this water-soluble polyethylene glycol spacer illustrated in the cartoon here at right with this blue ribbon, that peg chain actually helps protect the vaccine during its transit from the injection site to the draining lymph node. When we inject a free peptide, it'll be exposed to proteases and peptidases in the tissue immediately on injection, and that can lead to rapid degradation. If we, for example, in this experiment, take a melanoma antigen and incubate it with fluid from tissue to mimic this protease exposure, we see that the fresh antigen, which has a certain ability to stimulate a response from T cells, loses almost all of that potency after it's been incubated for 24 hours in tissue fluid.
An amplifier form of that antigen incubated in tissue fluid for the same time period shows no loss in potency because it turns out that polymer chain actually sterically protects the peptide, and proteases physically can't access the antigen as it's being protected by that polymer chain in the tissue. Next slide. Finally, a third feature of this albumin hitchhiking strategy is that the lipid tail has a second important role in the delivery, which is once it reaches the lymph node, these molecules can transfer off of albumin and insert their lipid tail into cell membranes in the densely packed structure of the lymph node, which is filled with lymphocytes sort of shoulder to shoulder.
When we inject these amplifier vaccines and then look in the lymph node at a day or two days or three days after injection, we can carry out cell sorting to take individual cells out of the lymph node that we see taking up these amplifier vaccines. We can detect that the amplifier is actually decorating the surfaces of these cells by staining with antibodies against the amplifier molecule. We see that the cells that take up the amplifier in their membranes are primarily antigen-presenting cells, macrophages, and dendritic cells, the exact cell types that you'd like to have accumulating this vaccine.
This plays at an important final step in the process, which is the albumin carries the vaccine to the lymph node and then transfer into cell membranes, makes sure that the vaccine gets captured in the lymph node, and it doesn't keep going on into the bloodstream. These three effects of better transport to lymph nodes, protection from degradation, and transfer into cell membranes changes are really important in the immunological property of the vaccine, which is the duration that antigen is being presented to T cells in the lymph node. When the vaccine gets taken up in the lymph node by dendritic cells, the peptide will be cleaved from the peptide from the amplifier chain inside dendritic cells and then processed, and the antigen will be presented to T cells in the collective MHC molecules on the surface of the dendritic cell to cognate T cells.
In this experiment, we asked how long is the antigen presentation continuing. We do that by vaccinating animals with either a soluble peptide vaccine or an AMP vaccine. At different time points post-vaccination, we inject transgenic T cells into the mice that all recognize the vaccine antigen. Those are a kind of biological reporter. Those T cells will traffic into lymph nodes in the animal, and if they see their antigen, they'll become activated. We can look one day after we transferred those T cells into the recipient mice and ask, did those reporter T cells get activated? That'll tell us that the vaccine is still being presented. When we look two days after immunization, as shown in the graph at right, both the peptide vaccine, the traditional peptide vaccine, and the AMP peptide are both stimulating T cells nicely.
If we look seven days after vaccination, now, because the traditional peptide vaccine got such a small amount of vaccine to the lymph node, it's completely cleared away by seven days, and there's no more T cell stimulation happening. The AMP vaccine is still going strong at that time point. When we've done this over time, we know that in the preclinical model, the vaccine is being presented for about two weeks. A dramatic enhancement in the duration of presentation from a few days to a few weeks. That translates into a dramatic increase in immunogenicity of the vaccine. How many T cells do we activate with this process?
Now if we go to a vaccine experiment where we vaccinate mice twice at day zero and day 14, and then at three weeks, we read out how many vaccine-specific T cells did we elicit. We can see in this experiment with a melanoma antigen, we go from a traditional peptide vaccine where about 2% of the CD8 cells in the animal are reacting to the vaccine. If we give the same dose of vaccine as an amplifier vaccine, now in some animals, as many as half of all the T cells in the mouse respond to the antigen.
In the graph at right, you're seeing this replicate data from groups of mice comparing either short 9-mer peptides or "long 20 amino acid peptide antigens." In both cases, the soluble vaccine elicits a very weak response that is magnified somewhere between 20-30-fold by using an AMP vaccine. Next slide. I'll just also note that it's just as important to get the adjuvant delivered to the lymph node efficiently. In experiments where we varied the amplifier structure to get better or worse delivery of a molecular adjuvant to the lymph nodes, in this case, Amphiphile CpG, we saw that the vaccine response was also sensitive to how well we delivered the adjuvant.
If we measure how much CpG adjuvant accumulated in the lymph node and then measure how much of a T cell response you elicit, you can see there's a direct correlation. If we deliver 30-fold more CpG to the lymph node, we got a 30-fold greater T cell response. This is also part of what Elicio is exploiting in their platform. To finish this one piece of preclinical data, now the question is, what does this do for the efficacy of a cancer vaccine? This is just one representative experiment in a head and neck cancer model where we're vaccinating against HPV antigens. If we give a traditional peptide vaccine in red, you see it slows tumor progression down slightly, but none of the animals will be long-term survivors.
In contrast, the same vaccine delivered as an amplifier form regresses tumors, and about 40% of these animals will completely reject the tumor even with a vaccine monotherapy. Next slide. To conclude, what we've seen is that the AMP technology enables efficient entry into lymphatics. It protects the antigen and adjuvant from premature degradation, and it promotes capture in the draining lymph node. Those three effects together really dramatically amplify T cell priming in preclinical models. That is what motivated us to launch Elicio and ask whether this could translate to humans. Here we'll hand off to Eileen to walk you through the clinical problem of PDAC and the problem of targeting KRAS. I'll stop there.
Great. Thank you very much, Darrell.
It has been a pleasure to work with the wealth of preclinical data to help to design the clinical program for translation of this technology. It is now my pleasure to welcome Dr. Eileen O'Reilly as the next speaker. She is the Winthrop Rockefeller Endowed Chair in Medical Oncology at MSK and also serves as the Section Head for hepatopancreaticobiliary and neuroendocrine cancers in the GI service. I will just highlight a couple of other sort of national roles that Eileen plays as the Co-chair of the NCI Alliance Cooperative Group, the service also on the NCI GI Cancer Steering Committee with the ASCO guidelines, and also as an associate editor for the Journal of Clinical Oncology. Welcome, Eileen. We are so happy that you were able to join us today to talk about adjuvant and neoadjuvant treatment in pancreatic cancer.
Thank you so much, Chris, and thanks, Darrell, for the great background for this. Delighted to be here this afternoon. Next slide. I do need to acknowledge various disclosures. All uncompensated for me. Thank you. If we could go to the next. I'm going to set the scene for pancreas cancer and provide the background for how we translate what has been discussed to the clinic. For those who may not be in the weeds on this disease, I think it's worth just reminding us all of the magnitude of this problem. It's a disease with a rising incidence. It incurs substantial morbidity and mortality and is in urgent need of new approaches. With that, we'll go to the next slide. The disease is primarily treated by chemotherapy, and this is in the advanced disease setting.
Almost all the benefit thus far has been with cytotoxic treatments. There are four regimens that are used in the first and second line that have shown the ability to shrink the cancer, control disease, and to extend life. As you can see, these numbers are overall relatively modest and have been incremental advances in the treatment. Not to say that they do not meaningfully impact for select individuals. They certainly do. Next slide, please. The approach to targeting pancreas cancer from a more selected perspective is really now a very active one in the clinic. This has been sort of underpinned by the observation that if you have a target and you match that target with a therapeutic, you can improve outcomes. This is a retrospective analysis of a community and academic-based genomic profiling cohort of individuals with advanced pancreas cancer.
This red line speaks to a group that had a target and a matched therapeutic. The two lower lines, those that either did not have a target or did not receive a matched therapeutic. You can see even with this retrospective analysis and all the constraints of that, that there was a striking signal in pancreas cancer in the advanced disease setting. That has sort of been the background for many years. This was in the pre-RAS therapeutic area and mostly here by some small subsets with BRCA-related malignancies. If you could go to the next slide, please. Moving forward, we will not focus on this in great detail, but just to say the biomarker-selected opportunities are extending in this disease. The critical one, the ubiquity of RAS, has not been able to be therapeutically exploited until recently.
There are other things happening in pancreas cancer as well, and even with regard to refining standard therapies, looking at transcriptomic subtypes. If you could go to the next slide, please. Now if we move to localized pancreas cancer and a reminder here for how the disease is approached clinically, that is a spectrum for those who do not have metastatic disease, for those on the left who have a resectable cancer where it is operable today, to those that have locally advanced disease who are not candidates for surgery but might be in the middle, the borderline resectable group, if they receive preoperative or neoadjuvant intent therapy, a percentage will be downstaged and able to undergo resection. This is a fairly significant subgroup. Just noting for the resectable population, the two approaches today, the traditional approach is surgery first, followed by postoperative adjuvant therapy.
Increasingly, there's been a shift toward neoadjuvant therapy promoting earlier delivery of systemic treatment in a disease where there's a high risk of systemic relapse. A series of studies have suggested that that might for select individuals be a preferred approach. Next slide, please. Looking at this, neoadjuvant therapy shrinks the disease. It can facilitate a margin negative resection. That's an R0 resection, which is oncologically desirable. It can select out those individuals who may be destined for early recurrence and would not be well served by a major intervention. Using treatment biology and time, neoadjuvant therapy can be very helpful in this disease. In the resectable population, it remains a point of discussion as to which is the best way to proceed, whether upfront surgery or preoperative therapy. Next slide, please.
A lot of information on this slide, but I'll just summarize it very briefly to say there are now multiple randomized trials in the adjuvant setting. For an individual who's undergone surgery and eligible for postoperative treatment, the early phase looked at chemotherapy versus observation. That was a win. The more recent phase looked at multi-agent combination chemotherapy against single-agent chemotherapy. There's been a gradual shift in outcome for combination chemotherapy with the biggest advent from the inclusion of FOLFIRINOX in the guidelines. If you could go to the next slide, these are the data looking at modified FOLFIRINOX, oxaliplatin, irinotecan, and 5-FU in the adjuvant setting compared to gemcitabine. Primary endpoint was disease-free survival and secondarily overall survival. You can see there's nice differences between these curves and these data held up on a more recent publication.
That's the go-to regimen in the adjuvant setting. It's also, for the most part, the go-to regimen in the neoadjuvant setting. There are two major studies that are anticipated to complete recruitment in 2025, one from Europe and one from the North American Cooperative Group System. Both are addressing this question of, for an average individual with a resectable, and I'll just specifically emphasize resectable pancreas cancer, whether a preoperative or perioperative approach is preferred over surgery first. It will take a few years to read out. That kind of summarizes where we are today. If you have an operable pancreas cancer, you can go to the operating room or you can receive preoperative treatment. If so, that's most likely to be FOLFIRINOX.
Similarly, if you're resected and of good functional status postoperatively, modified FOLFIRINOX is the recommendation, with gemcitabine and capecitabine being an alternative. Of course, clinical trials hold a high priority niche. Next slide, please. Now moving to KRAS-directed therapy. Next slide. Just again, a reminder of the significance and importance and magnitude of KRAS in pancreas cancer. This is sort of older data in a recent review. You can see that it's the disease of RAS and in malignancy in general, about 20% will have a RAS mutation and about 75% of those are in KRAS. In pancreas cancer, it's almost all in KRAS.
If you go to the next slide, please, and then as you look deeper, you're likely to find even more RAS in that almost everybody with this disease will have alterations in the RAS pathway or the MAP kinase pathway. A very small percentage, a couple of percent of people will have RAS wild-type disease, and they typically will have other actionable drivers. Next slide, please. There is a lot happening in this space, as the audience is likely more than aware. There are several KRAS targeted agents that are in the guidelines for pancreas cancer in the setting of KRAS G12C. There are small molecule inhibitors, noting that G12C is about 1-1.5% of people with pancreas cancer. Multiple ways of targeting RAS directly, indirectly, proximally in the pathway, downstream in the pathway, and of course, immunologically. If you could go to the next slide, please.
Thank you. One more. The pancreas cancer has been extraordinarily challenging for multiple reasons. The gain that the oncologic fields accrued from the Licensure of Immunotherapy has for the most part not been realized in pancreas cancer. There have been multiple approaches over 20 years and more looking at trying to make this disease immune responsive. Some of the reasons for these challenges are noted here, that it is a very immune suppressive microenvironment. It is a hypoxic, relatively avascular environment, at least of the primary malignancy, low number of activated T cells, a lot of immunosuppressive cytokines, relatively low number of neoepitopes, and low rates of mismatch repair deficiency, all of which can militate against mounting an effective immune response. Having said that, there are hints that for individual people, there can be striking and sometimes profound and durable outcomes to immunotherapy.
Part of what we heard as the introduction is, I think, the secret to sort of moving this field forward. If we could go to the next slide, please. This is an example of using a highly sophisticated approach with a T cell receptor therapy in pancreas cancer, showing that you can induce an immune response. Some of these immune responses can be relatively durable. Some of these TCR-based approaches are contingent on select HLA types. A lot of logistic considerations in terms of scalability and feasibility. Speaking to the time challenges of pancreas cancer, this can present some challenges in the clinic. Nonetheless, shows that effective immunologic select targeting can be a benefit in this disease. Next slide, please. With that, we'll move to the clinical development of the Elicio ELI-002, initially two peptide vaccine targeting G12D and G12R.
This was conducted in a unique setting in pancreas cancer. Those who had completed all of their standard therapy for localized disease or resected NED but had to have either rising biomarkers, EA099, or had to have detectable ctDNA. This was sort of a window of opportunity study looking at safety, looking at immunogenicity, and exploratory looking at clinical outcomes. A classic prime boost approach was used for this vaccine. From the administration part, this is actually very straightforward in the clinic, a series of weekly injections for four weeks at four different sites, every other week, a pause period, and then a boost phase for four weeks and follow up. Next slide, please. Highlighting the key issues, the study included both individuals with colon and pancreas cancer, primarily pancreas cancer.
This slide just illustrates a couple of things, that there was reduction in ctDNA and/or biomarkers for a majority of individuals. There was clearance of detectable ctDNA in about a quarter of individuals. It was a very well-tolerated approach, mild fatigue, skin reaction site, erythema, and discomfort and malaise. Relative to many things that we do, this was all grade one, maximum grade two. You can see as the dosing and in the initial phase one, it was the adjuvant that was increased rather than the peptide dose. There was clearly a clinical signal. If you could go to the next slide here. That clinical signal was when you looked at the median T cell response, those who were above the median versus below, you'll see that there was a nice segregation of outcomes.
To make the point that a group of individuals with rising biomarkers and detectable ctDNA, it is almost inevitable that recurrence will be evident in a relatively proximate timeline. This kind of signal suggests that something meaningful is happening. This has induced a CD4 and a CD8 T-cell response. Chris, in a moment, will talk more about the immunologic data. If you could go to the next slide. Looking at the two peptide vaccine, again, G12D and G12R, at the time of the initial data cursed and manuscript, the signal was very robust. A year later, you can see that signal holds up very nicely for those who had an immune response above the median T-cell threshold. More to come on this. If you could go to the next slide, please.
Looking at the colon and pancreas subgroups here, you'll see very favorable outcomes in relative terms for relapse-free survival, not mature yet, for overall survival for this group of individuals with a high risk of early disease reemergence. Next slide, please. Now moving to the seven peptide. This is all early data. This is covering all of the key amino acids that are relevant to KRAS in pancreas cancer. This used the same dose of the immune adjuvant at the recommended phase two from the two peptide trial, but used a higher dose and explored a higher dose of the peptide. Both appear to be important, the dose of the immune adjuvant and the higher dose of the peptide. Here, the magnitude of the T cell response was substantially greater. There were also similar observations with regard to the biomarker data.
You can see with this seven peptide vaccine, a very similar early signal with regard to segregation for those who were immune responders versus not. If you go to the next slide, please. That translated into this randomized phase two, the AMPLIFY-201 study that has completed recruitment. It was built off the 2P and seven peptide experiences. It included individuals resected NED, had to have an absolute lymphocyte count greater than one. That was one of the key observations from the phase one. This study is a two-to-one design in favor of the experimental arm with the possibility of crossover if certain criteria are met at the time of disease recurrence to establish if there is a clinical signal in that context as well. Primary endpoint here being disease-free survival. Also looking at biomarker clearance, other critical clinical endpoints, and immunogenicity. Next slide, please.
Just supporting the signal in a different way with a different platform is the experience that's led by our colleagues here at MSK looking at personalized neoantigen vaccination in pancreas cancer. This was a small phase one experience combining this vaccine with an immune checkpoint inhibitor, PD-L1 inhibitor, sequenced to chemotherapy, and then a boost phase for this personalized vaccine delivered in the postoperative setting. Once more, there was a segregation of responders and non-responders immunologically correlating with outcome. If you could go to the next slide, that has led to this study, which is underway also in the resected population of one-to-one randomization using this personalized neoantigen vaccine combined with immune checkpoint blockade and sequenced with chemotherapy and compared to standard of care. Next slide, please.
I'll sum up with two last slides here to say these are very exciting times in pancreas cancer, having been in the drug development field in this disease for a while. Potent early immunologic and clinical signal with the two peptide vaccine expanded with the seven peptide vaccine, acknowledging that's early, we need maturity. Lymph node targeting, as we heard eloquently from Darrell, expanded both CD4 and CD8 T cell response. These T cells are functional and active. There's cross-antigen reactivity. This was associated with a reduction in biomarkers, clearance of ctDNA, and delay in relapse for a percentage of individuals. We eagerly look forward to the readout of the randomized phase two.
If you could go to the next slide, I'll just present a couple of ideas as to where immunotherapy, and in particular, RAS immunotherapy, and these types of platform approaches can have value in pancreas cancer. Reverting back to the beginning of the talk, where the field delivers a lot of neoadjuvant therapy. This may be a really interesting space, just using the tumor as the adjuvant in situ, looking for combination data with immune checkpoint blockade, with chemotherapy, with targeted RAS inhibitors, looking at translating this to a debulked stage. You could think about advanced pancreas cancer as we get into the year of more effective therapy, minimal residual disease states, an ideal opportunity for integrating a low toxicity, highly immunologically active approach in this disease. Of course, this is only the start as we think about incorporating multiple epitopes.
If anything, one of the challenges for the field will be how to prioritize and which particular strategies to bring forward. I'll stop there and turn it over to Chris. Thank you all for listening.
Thank you very much, Eileen, for that wonderful overview of the adjuvant and neoadjuvant space, the current trial data, and the look towards the future as well. It's a very exciting time. I'll turn here to going over a little bit about the development program and the upcoming interim analysis just to remind everyone of what's coming up in the clinical development of the ELI-002 program. As you've seen from Eileen, there's some interesting preliminary data here suggesting the potential to change the treatment paradigm in this window of opportunity for the pancreatic cancer population.
Our observations in the phase one trials have shown that there was good tolerance at all the dose levels, no dose-limiting toxicities or serious side effects. The local reactogenicity has been manageable, often with cold compresses and many patients also needing to take over-the-counter medications such as Tylenol or nonsteroidals, for example. With that, we've established the phase two dose in the first two peptide trial with a 10 milligram amount of the AMP- CpG, as well as the higher amount of the amplified peptides. I think this development program has been really able to benefit from the systematic optimization of the biologic dose that could elicit the highest T cells. The T cell response in this case with the higher peptides was able to increase even an order of magnitude from what was published with the two peptide trial initially.
The ability to elicit the CD8, as Darrell nicely outlined, that has been such a goal of the cancer vaccine space was reflected by the high proportion of patients, about 60% with the two peptide data and even higher for the seven peptide version in the patients who received this treatment in the early studies. We did see the T cell response, as you saw from Eileen's talk, correlate to the anti-tumor effects, including the reduction in tumor biomarker levels. Another interesting phenomenon that we did not have time to cover in too much detail today, but antigen spreading is occurring where we can see bystanding T cells joining the RAS-specific cells. That may also be an important contributor to the anti-tumor effects here.
If we turn to the next slide, you can see that there's really a lot of potential for this therapy to move forward. Our first registration, we intend to be in pancreatic cancer, where, as you've seen today, there's such a high proportion of patients who harbor these driver RAS oncogene mutations. There are additional patients. We've got a handful of patients with colorectal cancer that have been studied with this agent. We haven't really begun to address the other populations yet, which are several. Overall, a quarter of cancer patients with solid tumors harbor a RAS mutation. About a quarter of lung patients have these, as well as smaller but significant proportions of biliary tumors, ovarian cancers, and others as well. This could be able to help patients with a variety of these driver oncogene mutations.
If we go to the next slide, we can see that the approaches that are coming that were nicely outlined include the small molecule therapies. Now, the first small molecules for RAS that have been approved target G12C. That's primarily relevant to tumor types like lung, where the G12C allele is more common. Additional inhibitors are coming for other alleles that may be helpful in pancreatic cancer. We've also seen that there are advances, as was outlined, with both personalized as well as the AMP off-the-shelf approach. We think that this is holding a lot of promise because of the ability to have that immediately available for patients because it's off-the-shelf, able to generate the immune responses with the key CD8 T cells and potentially could synergize, as was outlined for the future. It'll be very interesting to try combinations as the field progresses.
If we go to the next slide, I just wanted to highlight a little bit of the design features of the randomized phase two that Eileen introduced. Relative to the phase one studies, I wanted to point out that we made the population a little bit more uniform in consultation with FDA. Initial trials had allowed patients with stage four resected disease to enter. For the phase two, we've confined that to patients presenting with stages one through three disease, a complete R0 or R1 surgical resection. We did allow for patients with both MRD-positive tumors, so patients with circulating tumor DNA or progressive protein, serum protein levels following their local regional treatment to participate.
The trial is two-to-one randomized. The patients receive either the ELI-002 or are observed, which is the current standard for the window of opportunity, with the primary endpoint of disease-free survival and the other important secondary endpoints that you can see listed here. Just in the interest of time, I'll just mention in brief that we've met with the FDA about the design for the phase three that would support the registration here and have determined that the disease-free survival endpoint is expected to be the primary endpoint also for phase three. The trial would have a very similar design, would be randomized, and in this case, for phase three, blinded. We'll continue to use the investigator's assessment of disease-free survival and continue to confirm new lesions either by biopsy or imaging. Let's go to the next slide.
At a very high level, what will set the stage for the interim analysis, the trial has been designed to reduce risk by having some features like the stratification by node status. By making sure that the randomization balances those patients with a node-negative versus node-positive pathology at their surgery, we will avoid imbalance in the prognostic factors. We used a weighted average of the disease-free survival expectations for the MRD-positive and negative groups with a high power for the study. What we expect the IDMC to review will be the primary endpoint, the disease-free survival, and the important secondary endpoints. The overall survival will be of interest as well, just reminding everyone that crossover is permitted in the trial so that could affect the overall survival. The overall survival and the one-year DFS rate may require a little bit of time to mature.
I wanted to set the stage that if the trial is declared a success by the IDMC, we may have to wait a little bit longer for some of the secondary endpoints to mature as well. Of course, the interim data could lead the IDMC to recommend to consider that to be the final analysis. We could be advised to continue to the final analysis, or it could be determined that it's a futile data outcome. Of course, we don't think that's as likely based on the phase one observations with this agent so far. Okay, let me turn to the next slide then. I'll open the, just in the interest of time, I'll just move to open up to Q&A discussion here. Great. Thanks, Chris.
At this time, we'll be conducting a Q&A session with our speakers.
Please hold for a brief moment while we pull for questions. Our first question comes from Soumit Roy at Jones Trading. Please go ahead, Soumit.
Good afternoon, everyone. Thank you again to all the panelists for comprehensively presenting the landscape. One question on the endpoint and the FDA, is this going to fall under CDER or under CBER? Have you had a recent contact and discussion saying this is still the go-forward FDS can be used for the phase two and phase three registration endpoint?
Yeah, thank you. The file for ELI-002 is being reviewed by the CBER group. There have been some personnel changes there, but we are not aware of any changes. We will conduct an end-of-phase two meeting before we initiate the phase three trial. That will be with the data in hand from the randomized phase two that we have just discussed.
We will have another opportunity to make sure that we maintain alignment with the FDA as we proceed with the development program. Okay. The second question is on the, so you're allowing the crossover in the phase two. Could we expect that data because that would be the first indication, first scenario where we will get progressive patients undergoing with the drug? Could we expect that data around the top-line data that we're expecting in third quarter, or are you going to separate those two events? Oh, it may take a little bit longer for that data as patients will only begin that treatment at the time that they progress. The start time for the patients who are electing to crossover is a little bit later, and the data is less mature. Right. Thank you again for taking the questions.
Great. Thanks for the question, Soumit.
Chris, we can go to the writing questions from here.
Okay, great. I think I'll just do one more question maybe for Dr. O'Reilly. A couple of questions here came as you've outlined so nicely, the chemotherapy being the mainstay of treatment so far in the adjuvant and neoadjuvant settings. With these novel therapies coming along, do you foresee a future where it might be possible to eliminate chemotherapy in pancreatic cancer?
Yes, thank you for that. It's a great question. I think we can actually see that there's a horizon on this, which hasn't been present in this disease. I'm speaking about the totality of novel therapeutics. For most people, it's fair to say, I think for the immediate future, cytotoxic therapy is going to be a critical component of their treatments.
It is likely for some groups of people that in the not-too-distant future, there will be opportunities for non-chemotherapy, either single-agent or combination approaches. It's very exciting to think about.
Great. Thank you so much. I think we're close to the time here. Let me pass it back to Bob Conley.
Thanks, Chris. Let me really thank Eileen and Darrell and Chris for their time today and the effort that they put into the presentation. I think you can see that we're very excited. The phase one data, the enthusiasm of our investigators and our patients for what we're doing really gives us a lot of encouragement here as we move into the planned Q3 initial analysis of our initial randomized controlled study.
With success there, it's a big step towards these PDAC patients who so badly need this type of approach and this type of success. I think it also for us moves us more towards expanding what we do at the ELI-002 into perhaps neoadjuvant, as Dr. O'Reilly was talking about, but also into other KRAS-driven populations like colorectal cancer and lung, and even into combinations where we believe that combining our approach with other approaches, whether they're small molecule or immunotherapies, could have a dramatic impact on those therapies as well. I think just a general feeling here that we've just scratched the surface with this program, as well as with lymph node targeting and the impact that that can have in the oncology world. Thanks for everybody. To everybody for joining. Stay tuned. There's a lot more to come from this company.