Again, thanks for having us. Michelle and I are now going to dive into an explanation and review with you our AACR three presentations that the team presented at the meeting in Chicago just late last month. Let me quickly bring up the slides, and we can go forward then. Michelle, thanks again ever so much for being here.
Pleasure to be here, Chris.
Let's dive in. The AACR, the first presentation that we're going to look at was by Rory Edwards, who is our Head of Translational Medicine. It was titled, "Investigating Fibroblast Activation Protein as a Therapeutic Target for Delivery of Precision Cancer Medicines: Expression, Spatial Localization, and the Functional Insights." I'm just going to show you a couple of the key figures that were in this particular presentation. As you can see, this is our usual schematic and going through this concept of the bystander effect. What do we mean by the bystander effect? What it means is that the drug is cleaved. Recall, Avacta uses a peptide drug conjugate, and the peptide, when it is attached, you can see the attached peptide drug conjugate modeled here. Here is the active payload or the cancer drug, and here in white is the payload.
What happens with the bystander effect is that the peptide itself is removed by the actions of the FAP enzyme. What happens now is that the peptide drug conjugate, when it's conjugated, it can't enter the cells. It then gets released, but it is released by an enzyme, FAP, that is expressed on the cancer cell fibroblasts, not on the tumor cells in the vast majority of human tumors. The peptide is released. What happens then is that we now have released active free payload. It can now move into and be taken up into either a FAP negative tumor cell or a FAP positive cat. It is able to kill both of those cells. Critically important because the vast majority of human solid tumors don't express FAP on their surface. That released payload now moves into the FAP negative tumor cells.
Before we got to this particular study in the poster, we looked at the RNA versus the protein. If you'll recall, the way that a cell makes a protein is that it's encoded in the genes in the DNA. That DNA is then what we call transcribed into the messenger RNA. And that RNA is what goes out of the nucleus and moves there and essentially is the blueprint for the manufacturing of the protein. To study the protein level requires tremendous methods, and it's easier. In fact, we can study many, many, many human tumors at an RNA level, much easier to study the nucleic acids. What you're looking at here is a large study and our strategic collaboration with a company known as Tempus.
Within the Tempus database, there are 160,000 human solid tumors, and they've been profiled at many levels, including at the RNA level. What we showed just before this in the poster is that the RNA level is highly predictive of the protein level. We can use these kinds of databases to really do a very large study of the addressable market or the market opportunity for precision medicines. Now, what's important about this one is that this figure doesn't assume that it's a single medicine. Recall, one of the important things that we found or one of the important findings in the AVA6000 trial is that it is the warhead or the payload sensitivity that is important. Recall, doxorubicin is used in three solid tumors. It's used across the board in breast cancer. It's used in soft tissue sarcoma and in combination in salivary gland cancer.
Let's look at a couple of the key points of this particular slide and a couple of the key market opportunities that we're interested in. This is both those disease settings that we will be looking at with AVA6000. There you can see triple negative breast cancer. The global incidence is 170,000. FAP positive is 161,000. Very large market opportunity. Metastatic salivary gland cancer is about 15,000 globally, with FAP positive being 13,500. Now, if you look at the positive rate there, we won't need a companion diagnostic in either of these particular disease settings, which allows us to move as fast as possible towards these potential registration trials that we will look at. As we move towards AVA6103, or our next entry into the clinic, AVA FAP-exatecan, we're interested in some of the GI cancers where TOPO1 mechanism is quite relevant.
Pancreatic ductal cancer, 495,000 patients globally, most of which will be FAP positive. Gastric cancer, another critically important one where we know two different lines of evidence TOPO1 is quite active there. FAP high, 400,000; FAP positive, 850,000. Small cell lung cancer, another place where we know TOPO1 as a single agent works well. Topo TCAN is used. There is monotherapy. FAP positive, 200,000. A large addressable market opportunity. One of the questions that we ask ourselves, especially as we move into some of these more stroma-heavy tumor types, triple negative and pancreatic cancer are probably the two that come to mind, is where is the FAP expressed in relation to the tumor cells? Because really, we would like the drug to be primarily released right at the tumor stroma interface.
One of our scientists, Curtis Rink, went through this exercise of essentially taking a stain of a human tumor and breaking it down and allowing the computer to read concentric or making circles essentially around the tumor cells and asking the question, what happens to the FAP expression as you move further away from the tumor cells? Exactly as we planned, you can look down here and see regions one through four, which are closer and furthest away from the tumor cells. What we see here, critically important, is that most of the FAP is expressed right next to the tumor cells. The drug will be released right in the setting where we need it. That was our first poster.
The key findings there, the bystander effect, FAP expression in the addressable market population, and critically important that we are able to see FAP right at that tumor stroma interface. Let's now move into our second presentation, which now brings us to the update on the clinical data with AVA6000 or FAP-dox. This poster presented by Dr. Lau Gu, Comparative Pharmacokinetics and Tumor Activation, Fibroblast Activation Protein Enabled Precision Peptide Drug Conjugates. We looked at a couple of different ones here, both FAP-dox as well as FAP-exatecan. We will have seen this slide before, but this is the phase I trial of AVA6000. You can see the different dose levels there. We escalated through both dosing regimens, every three weeks and then every two weeks, and did not identify an MTD in this trial.
We will be taking forward a recommended dose for expansion, looking across all of the data collected at the various higher doses that you can see there. We analyzed for efficacy. First, let's look at the safety. This is a summary slide just looking at some of the key safety findings out of the trial. We did present the safety by dose level in the poster. To summarize, bone marrow toxicities, when we use AVA6000 and compare it to conventional doxorubicin, are dramatically reduced. Just as an example, half of the patients that receive the conventional dose of doxorubicin, 75 milligrams per meter squared, will have severe neutropenia, and that is significantly reduced across dose levels with AVA6000.
Importantly for us, in the choice of doxorubicin and being that key first drug to FAP enable, one of the key points here is that there is a toxicity that we knew if we could eliminate it, we would have proof of concept of the precision platform. That key toxicity is the cardiac toxicity. What we see with doxorubicin as we dose up to that lifetime maximum, be it 450 as described in the label versus 550 where we now take these patients, we can see out to 550, as you can see in the figure, that 6 to 20% of patients will experience a severe cardiac toxicity. Importantly, with AVA6000, despite most of the patients receiving doses above that 75, AVA6000, zero. We think of that really in terms of the safety data as proof of concept for the precision platform.
We've been able to eliminate that severe cardiac toxicity. Important for patients is that alopecia is generally limited to hair thinning, and there are no reports of ADC-related toxicities. These toxicities, pneumonitis, ocular toxicity, liver toxicity, are associated with a nonspecific release of the warhead. One of the key advantages to the precision platform is that the payload, the warhead, is only going to be released in the tumor. All of this is fantastic as long as we're seeing activity. Let's look at that specific disease setting where we have seen some exciting initial data, salivary gland cancer. It isn't a smaller indication. It is one that we will be moving forward with. This is the waterfall plot. What you can see is that these are all patients treated in the phase I. This is not the expansion cohort data.
Each bar in the waterfall plot represents a single patient. It represents either tumor growth or tumor shrinkage and what is the best response that happened. Importantly, only one patient had as their best response tumor growth that qualified for progression. What that leads us with is that 10 of the 11 patients in this cohort, in this group, had either stable disease or a partial response or a minor response as their best response, meaning the disease control rate here is very exciting, 10 out of 11, over 90% of patients had disease control. You can see there are three patients here that are still on study with a small degree of tumor shrinkage, four patients here that are still being assessed for PFS. They have reached their lifetime maximum of AVA6000, four minor responses, and one partial response. Multiple patients with tumor shrinkage here.
What matters, though, to patients and to us as drug developers is how long are these durable tumor shrinkages? The answer, although we haven't reached median progression-free survival yet, is yes, these are durable. The arrows indicate in this swim lane plot, which plots each patient over time on trial and their weeks of treatment and follow-up, these arrows indicate those patients that we are still following for progression-free survival. The black dots indicate patients that are still in follow-up. The triangles indicate the scans that have been done on the study. What's important here is that because of the number of the aqua arrows, because eight of the 11 patients are still in PFS follow-up, we haven't reached the median.
When we compare the benchmark, which is 3.5 months, that was published at ESMO last year, a large series by the EORTC in salivary gland, we have nearly doubled the PFS with our median follow-up now 5.9 months. Because we haven't reached median yet, this number can only get better. We're very excited about these data showing that it is very likely that we will be able to beat those historic controls that we see with the published data and will be moving in the first quarter of next year towards that phase two, three trial of AVA6000 and metastatic salivary gland cancer. This is my favorite figure in the poster. The reason is that it's a very simple experiment, but it shows the power of precision.
What you can see here on the left is a simple experiment in a number of the patients where we were able to obtain a biopsy 24 hours after that first dose was taken. We see a two log concentration when we compare the tumor concentration to the plasma concentration. We often get asked the question, how do we compare then to conventional doxorubicin? In a publication from 1990, what we show is that the median here is one. The concentration that we see in the bloodstream is the same as the concentration that we would see essentially in the tumor with no concentration of the drug without the precision mechanism. We see a two log increase when we move to the precision mechanism. As we dive into some of the deeper pharmacology that was published, here we're looking at a physiologic-based pharmacokinetic model.
This kind of pharmacokinetic model takes multiple inputs. You can see here for this PBPK model, what we used is both animal data as well as human data using conventional doxorubicin and human data using AVA6000. Three inputs. What we are modeling is the concentration in a particular cell, which is the cardiac myocyte. You can see on the right-hand side the output of the model using both human data with conventional doxorubicin in the red dashed line, and AVA6000 released doxorubicin as we measure. We are modeling here what happens in terms of, I am going to move, but what happens in terms of the two factors that cause the cardiac toxicity, which is topoisomerase II inhibition and reactive oxygen species.
You can see with the first pass effect, which I'll describe in the next slide, that conventional doxorubicin bathes the tissues, and we see that the concentration in the cardiac myocytes is quite high. It's the same as that concentration in the blood, very different than what we see with AVA6000, in that the concentration in the tissues is much lower because of the first pass effect, meaning that released doxorubicin leaks out of the tumor and moves into the tissues. What we see is a much lower concentration across the different dose levels of released doxorubicin versus conventional doxorubicin, explaining why, because we're not hitting these enzymes, essentially why we're not seeing that cardiac toxicity. Combining this model with our first pass effect, which is depicted here, first pass meaning what happens within those first minutes after the drug is infused into the patient.
Within the first few minutes with AVA6000, it goes to the tumor, and the red is depicting doxorubicin, meaning the doxorubicin in that first pass is only in the tumor leaking out again into the bloodstream, and then eventually a small degree of it into the tissues. Very different than conventional doxorubicin, where the tissues are essentially bathed immediately. It's been published that the distribution half-life of doxorubicin is only five minutes. So within five minutes, those tissues get bathed with active doxorubicin, hence then getting large concentrations. AVA6000 is going to have delays and lower concentrations when we move into the clinic or when we move into the bloodstream. The conclusions from the poster, FAP docs is safe and well tolerated, both dosing regimens. We've seen that in previous data published.
No MTD was determined up to a dose of 385, which is almost four times the dose. The disease control rate in salivary gland cancer being 91% with multiple minor responses, one confirmed partial response, and over 25 weeks of follow-up, nearly doubling the benchmark PFS. Despite that dosing, we are getting the concentration of 100 to one, and we see that the normal tissue exposure is much lower than with conventional doxorubicin. We also see in the PK here that we have the opportunity for a sustained release mechanism to be delivered. That is what we are going to look at next in our second poster. Very happy about the cardiac toxicity, but also very much so the salivary gland data as it's coming together. Let's move now to our third poster.
We have a third drug or a second drug in the pipeline, a third poster that was presented at AACR last month on AVA6103 or FAP-exatecan, FAP-EXD. This was published by one of our scientists, the scientist leading this program, Curtis Rink, and the novel peptide drug conjugate AVA6103 is a FAP-enabled precision medicine, a TOPO1. Why did we choose exatecan? Let me take you through this, and then Michelle is going to take you through some of the exquisite science presented here. Exatecan is the most potent of TOPO1 inhibitors. It has single-agent activity in a number of different disease settings in a bunch of phase two trials.
This was originally a Daiichi Sankyo drug, and it was delivered and administered to close to 1,000 patients across all of the phase I, phase II, and a single phase III trial that were completed here. What we can see is that it has very similar potency to a cousin, deruxtecan, which has been a very highly successful ADC warhead. Probably the most notable ADC is in HER2, active in both breast cancer as well as in gastric cancer. Importantly, you'll recall FAP is expressed across the board in both gastric cancer and in breast cancer. It's not going to be limited in these disease settings by the expression of, say, HER2, where we see in HER2. Finally, the third reason is that exatecan failed for two reasons that we believe the precision technology can fix.
The first is it has a limited therapeutic index, meaning the toxicity was just too great. We had to keep reducing the dose when exatecan was delivered as monotherapy in the clinic. It has a significant PK issue. It has a very short half-life. In contrast to doxorubicin, which has a long half-life of 35 hours, we needed to fix the half-life of TOPO1 inhibitor and take that nine-hour half-life out to at least a few days. We needed to get inhibition days of TOPO1. we believed at the outset when we took this program forward that we were going to be able to do that using some of our advances in chemistry, which Michelle is going to take you through now. Over to you, Michelle.
Thanks, Chris.
Yeah, what I'm going to show you here is a model of some of the precision molecules sitting in the FAP docking site. On the left-hand side, we can see the AVA6000, the doxorubicin molecule, which is a simple attachment to the precision peptide. What the image on the right describes is the advances that have been made in chemistry to create AVA6103 to both extend the plasma half-life, so make the drug persist in the body for longer, but also to allow us to tune the rate at which it is cleaved by FAP in the tumor. This way, we can control how long it's in the bloodstream, but also how quickly it's released in the tumor.
What's shown behind these blue blobs, labeled A and B, are the chemical structures that we've not yet disclosed, which represent two inventions that essentially advance the technology, but also create AVA6103. The aim is to have a sustained release delivery in the tumor with very limited systemic exposure of exatecan, which is optimal for this highly potent payload. The next slide shows some of the data that we've generated with this molecule. This assay is human breast cancer cell lines grown in three dimension in spheroids, co-cultured with human breast fibroblasts, so CAFs. What we show here is that in the absence of any of CAFs, so if we just have tumor on the dark blue line, we can see that AVA6103 is inert. When there's no FAP, there's really no activity.
What we observe when we add the fibroblasts into this culture is a killing of those cells. Those cells die, and it's a very similar potency compared to just normal exatecan, which is shown there in the peach-colored line. These data indicate that AVA6103 is doing exactly what we want it to do. In the absence of FAP, it's inactive. When we have FAP expressed on the fibroblasts, the FAP is cleaved and the warhead is released, and exatecan can enter the cells and result in potent cell killing. The next slide shows some of the in vivo data. These are human tumors that have been grown in mice. We call these PDX models. This is a gastric cancer model. In this model, we demonstrate complete responses with AVA6103 in this model.
Importantly, if I can show you the little circular picture there, which is essentially a slice of these human PDX tumors, you can see that the FAP in the brown is just on the fibroblast, and it's sort of layered throughout the tumor. The little graph beneath shows the response of this tumor model to a less potent TOPO1 inhibitor. This is irinotecan, which is currently used in the clinic. This is a really hard-to-treat model. Exatecan in the peach line, you can see, has some moderate responses, but when we dose AVA6103, we've really got this very potent anti-tumor effect. These responses are durable, so the tumors do not grow back. You can see again on the right-hand side, really just the depth of the responses that we're seeing. In some animals, complete killing of the tumor.
The next slide shows why we believe that this molecule really, really works. As Chris described before, when we set out to design this molecule, we were looking to extend this half-life in the tumor out to a number of days. This graph here shows some PK modeling. This is taken from data where we've administered mice with tumors with AVA6103, and then we measure the level of exatecan that's released at different time points during the experiment. We look in the tumor and we look in the blood. What we can see is within the tumor, we have elevated levels of exatecan released, and this is extended out for a number of days, so around 60 hours in this case, at levels that are more than sufficient to kill cancer cells.
That's the IC50. Pancreatic cancer is the amount of drug you need to kill those cells. This really explains why we're seeing those effects in those mouse tumors, and we've seen those effects in a number of different models as well, and allows us to really see the chemistry and the inventions playing out in these studies. The next slide shows our plans. We're now very excited to be moving this forward towards clinical development. We selected a candidate around the end of last year, and we're now conducting the manufacturing process and the toxicology studies required by the regulatory agencies to allow us to start a first-in-human clinical trial. The IND, so the application for this, will be submitted at the end of this year with a clinical trial starting in the first quarter of next year.
Very excited to see these data emerge and really looking forward to moving this forward into the clinic next year. To conclude what we've told you about AVA6103, it's a novel precision peptide. It delivers exatecan, a very potent killer of cancer cells, directly to the tumor at the tumor stromal interface. It's only cleaved in the presence of FAP on the surface of the CAFs, and this gives us that specificity for the tumor. We're killing cells via this bystander effect. The cells don't need to express a particular antigen, as you would expect for an ADC. It can kill any cells in its proximity. The kinetics are optimal as a result of these two key inventions our chemists have made to the capping group and the linker.
Of course, this results in these high levels of AVA6103 that we see in the tumors, leading to these durable anti-tumor responses in the mice. We were very pleased to be able to present this at AACR last month.
Great. Thanks, Michelle. Very much appreciated. Let's bring Katie back and move into our question and answer part of today's webinar. Katie.
Thank you both very much to Michelle and Chris there. We are joined by Barry Gibb as well from our team here at Turner Pope. As I mentioned at the beginning, lots of questions were submitted to us. Thank you very much indeed for taking the trouble and the time with these. A lot of them are very detailed. What we have done is some of them were similar. They had similar themes.
We've clubbed them together, and hopefully, you'll all get the answers that you want. We've basically put them into three sections. We've got firstly the platform and the company strategy, and then we're going to move on and we'll concentrate on those questions on the specific on the AVA6000 and AVA6103. Let's get stuck in. We're going to start with the platform and strategy, as I say. This investor here would like to know, thinking about the company strategy, what is the priority in achieving value for investors focusing on clinical trials and commercial opportunities with the existing pipeline of assets or other potential warheads? How does this align with clinical trial timelines over the next two years?
Thanks, Katie. Appreciate that one. The company strategy, let me describe this in three different ways. We do have our existing pipelines.
AVA6000 is deep in its clinical development now, enrolling its expansion cohorts in three different disease settings where we know doxorubicin is active. AVA6103, which is our next asset moving into the clinic, and we know that that one should be moving into clinical development starting in the first quarter of next year. Also, we have a fantastic group of scientists, which many of you have met through our R&D Spotlight series, who are working on other potential warheads. There, we're excited. We're not at the point yet that we can disclose any of those that are sort of next and coming in the pipeline.
I would say that over the next two years, the important catalyst for the company, as you know, would be the data in the expansion cohorts with AVA6000, the phase I data for AVA6103, and then also seeing that pipeline fill out. We think that the opportunities with the peptide, with the peptide drug conjugates, with precision, there are many of them, but you'll recall the IP around this new mechanism, this sustained release mechanism, was not filed until September, October of last year. That is really recent. Over the next two years, I would say we should be looking for all three of those coming to fruition.
Okay, let's move on to the next question then, shall we?
This person, this supporter, this investor says, "Would it make sense for the company to gain approval for the precision platform, proving it has no adverse effects and is inert to the patient, proving it's just a delivery mechanism? Then, when used in conjunction with an already FDA-approved drug warhead, such as DOX, it can automatically be approved for each warhead."
It's a question that we see quite frequently. Why don't we just get approval for precision alone? Precision, if we boil it down to its essence, it's a small peptide that is then attached to the drug. I often use my phone. This is the drug, and then here is the peptide. This is the drug. It's not that we can get the peptide approved and put it onto anything. It's not how drugs get approved.
With each iteration or with each different molecule that we create, we will go through the approval process. It will be specific trials that we will do to show that when the peptide is attached to this specific drug, that it moves there. Because they are then one molecular entity, that is the approval process. Each of them does have to go through their own individual approval process.
Okay. This person is asking, "Can you describe the Tempus collaboration and its benefits? Does the collaboration with Tempus AI give them any rights to a share in the commercial benefits of discoveries made using their dataset?"
The Tempus collaboration is essentially allowing us to use AI in designing trials and really to throw some of these AI benefits and big datasets at the design.
It is going to allow us to move from in AVA6000, it took us quite some time and empiric work to figure out what were the diseases that we wanted to go after. Instead, with 6103, we are going to use AI to really do that empiric work for us, to really shorten the timelines. I can take the second part of that question and know this is a strategic collaboration. It is essentially we are accessing the data scientists and the data at Tempus, but there is no commercial gain for Tempus in this. Michelle, maybe since your group really led this, can you give a little bit of color as to the Tempus collaboration and its benefits?
Yeah, thanks, Chris. We are very excited about the Tempus collaboration. I think the power of their dataset is just the sheer size of it.
We have mined 160,000 patient samples to try and understand the expression of FAP. This is the substrate for our precision peptide and other characteristics of those tumors. Not only do they have a very large database, it is also very well annotated. We understand the patient's journey from those samples. Sometimes we have multiple samples per patient. This allows us to ask key translational questions such as to address things such as who might respond to one of our payloads where FAP is expressed. We have used this on AVA6103 to identify some patient populations that we want to move forward with in our phase one clinical trials. It is really very powerful from a translational perspective to understand the science in the patient population we will ultimately treat.
Thank you both for that one. This person wants to know about the partnering strategy.
They're saying, "Could you please confirm what the strategy will be for partnering during 2025 to 2026?"
I'm going to answer this one in three different ways and give our sort of the three pillars of the BD strategy. The first is with AVA6000. We have been very upfront that for this particular asset, especially as it moves into some of those later trials, phase II trials, which are in the plans for 2026, we are looking for a partner. We don't have any specific guidelines around that, and we are having a number of conversations there. Many folks have been very interested in seeing some of the key data, as I described with that first one, that will be coming, clinical data that will be coming with AVA6000.
We have actually filed that new IP, and this is around the sustained release mechanism of taking precision and moving it into that sustained release within the tumor microenvironment. Now, that IP was filed just at the end of last year, so it's quite early. We did actually just present at the end of last month in April, sort of the full pharmacology package of the first asset around that sustained release mechanism, and that's AVA6103 that Michelle just took you through. And those data really unlock some of the conversations around that whole idea of additional warheads. Now, we've been asked a number of times, "What is the right time for a partnership or some kind of deal around 6103, the beast, as I call it?" This one is, it's not now. Once an asset has clinical data, that's when you start to think about that there.
Really three different ways of thinking about it, and conversations are ongoing there.
O kay. Now, this is the last question on the strategy part, and it's a very interesting one because they're asking about veterinary oncology. Avacta's precision platform has earned praise for targeting human cancers with precision and fewer side effects, which is commendable. With a technology that could clearly bridge human and veterinary oncology and with pet owners desperate for real treatments and not just tests, what is holding Avacta back from using veterinary oncology to validate the platform?
Let me answer this in a couple of different ways, and I'm going to actually ask our resident scientist to comment as well. Way back in the day when I was at University of Pennsylvania, I did something very similar to this in that we had a new technology.
We were looking at loaded B-cell vaccines, and we did do a companion animal trial. It was in the most frequent tumor that dogs get, which is canine lymphoma. Believe it or not, solid tumors in animals, specifically in dogs, so let's take the K9 variety, are much less frequent than lymphoma. We know that FAP is not expressed in hematologic malignancies. Really, the patient population would be the main tumor that dogs get would not be susceptible. That brings up a whole level of new science. Michelle, I don't know, do you want to comment on this one as well?
Yeah, no, it's a really interesting question, I suppose. In theory, many of the drugs we develop for humans could be applicable for animals and pets as well. We're not veterinary medicine experts at Avacta, so it's not something we've really explored.
We haven't particularly looked at any tumors from dog to see if they're FAP positive at this stage, but it is something that could potentially be done, but not an avenue we're actively exploring at this stage. Most of our animal models, all of our preclinical models, are in the mouse, and that's because there's a lot of really well-characterized tools there that we understand some translatability for those. It's not something we've done right now, but I think it's a very interesting question.
Yeah, it definitely is.
All right, let's move on now because we've got lots on the AVA6000. We have selected a few of these, and hopefully, as we say, everyone will get their question answered somehow. We're going to move on to the science. The data is exceptional, and every time it is updated, it gets better.
What, in your opinion, is the scientific risk left that could prevent AVA6000 coming to market, and what do you estimate are the chances of this drug being a success?
The risks, especially coming out of a phase I trial, which is where we are right now, the population that is generally treated in a phase I study is pretty heterogeneous. The patients are somewhat different from each other, pretty heavily pretreated. The fact that we are seeing some interesting activity in salivary gland cancer is very exciting to us, especially given the unmet need in that disease setting. The risk is always when you move across lines of therapy. We are taking pretreated patients, and the phase II trial will be looking at likely a first and a second line setting. That is one of the risks. It is also then, are we going to be able to enroll this?
That is one of the risks that we take forward with this. For both of those, given what we've seen in phase I and given the excitement around this drug from the investigators that are seeing these patients, we're quite excited about that. When you move from phase I, though, into a phase II and a phase III setting, we have to then bring in the comparator arm, and the drug actually has to beat standard of care. That is always the big risk. At this point, estimating the chance of success, we're going to build those trials and build in the statistics around the data that we see coming out of the expansion cohort. Everything that we do in drug development is designed to de-risk the next step.
The expansion cohorts are designed to really help us hone in on, is this the right patient population? Have we selected the right lines of therapy? What is the difference between the benchmarking data and what are the data that we're seeing in the expansion cohort? That's going to drive the size of the trial, the statistics. We do everything that we can to de-risk it. Yes, moving forward and very excited to move to the next stage.
The next one is quite specific, actually. It says, "Why has 250 mg, milligrams been chosen for efficacy data for the SGC? Did the doses below 250 show no meaningful efficacy?"
Interesting question. We had to make a cut in the data somewhere so that we weren't taking 80, 160.
It was not an easy decision, but I can tell you that in the data, especially in the PK data, there was a nice break between the lower doses and then the 250 and above. We have not selected the 250 going forward. What we are looking at in terms of analyzing the efficacy in the trial is 250, 310, 385. All three of those dose levels were grouped together in terms of looking at the salivary gland cancer. That was really a PK and a safety decision that was made. We will be looking specifically at the recommended dose for expansion as we move forward in the next stage of the trial.
The next one is quite detailed, so I take my time with this one. The recent RNS talks to a near doubling of the PFS in the SGC patients to date.
Can you please confirm the disease control rate and provide more color to the trend which sits beneath the PFS? For example, do those patients who are now off the drug and under observation continue not to experience disease progression?
Great question. This looks at two different efficacy endpoints here. One we call the response rate, and this is where we measure the diameters of the tumors. That is how we calculate this disease control rate. What the disease control rate is looking at is how many patients experienced some level of benefit, meaning stable disease, minor response, partial response, versus how many patients had progression as their first assessment. It was the first indication to us as we were running the phase I that something appears to be happening there and that only one of those 11 patients had disease progression as their first outcome.
That is why the disease control rate is 10 out of 11 or over 90%. What we need to assess is how durable are these benefits, meaning stable disease, minor response, partial response, because the durability is what is going to really matter for the patients. It is very rare for a patient. I was an oncologist, and when I practiced, it is very rare for a patient to really worry about, "What was the diameter of my tumor?" Really, what they want to know is, "Doc, how long do I have?" It is that durability. How long will this treatment last? How long will I feel good? How long before the tumor starts to grow again? That is the measure of PFS. It is what we call a time-to-event analysis. Where we are right now is looking for the median PFS.
Among those 11 patients, when six of them have progressed, we will now see that median because the sixth patient really defines what the median PFS is. We know from ESMO of last year that in pretreated patients, this was a series of 54 patients, that the median PFS there was 3.5 months. Now, our median follow-up, because we still have eight patients that are either on drug or in PFS follow-up, we still need some of those patients to progress before we reach that median PFS. Now our follow-up is at 5.9 months. That is what we mean by this near doubling. 3.5, one more month, and we will have doubled the benchmark PFS.
These are the kinds of statistics that we're going to be using to both plan and size and pick the number of patients in that phase two trial, phase two three trial that we'll be looking at starting early next year. These are the data that we're going to use to have the conversations with the health authorities around. This is the plans. We call it the end of phase I, but these are the plans that we're going to take forward next year as we move into those. Two different efficacy endpoints, really looking at the diameters, but then looking at the durability. It's important that we see benefit in both of those. It really shows the congruence of the clinical data.
This last one is actually asking about the timeline.
Given the volume of clinical data collected to date, the favorable safety profile and the encouraging tumor responses reported in the AVA6000 trial, all while delivering an already approved chemotherapy agent in a more targeted and tolerable way. This, as I say, they're asking about the timeline. They're saying, "Are you able to provide an expected timeline for the remainder of the phase I A, B program?"
Yeah. The phase I A, the dose escalation has completed. We will be reporting a little bit later this year the final data from that particular study. With that, we're going to be looking at sort of that long-term cardiac safety data. Not just right up to the end, but really following those patients afterwards. The phase I B is enrolling right now. We are enrolling in the various expansion cohorts. We are looking at data from those.
The salivary gland is the first one that we'll read out. The PFS there will be a bit shorter, and that will be later this year. We'll be able to take a first look at the salivary gland data. The breast cancer data will probably take us into early next year. In the corporate deck, we've been very upfront about the timelines of a potential phase II or phase II, III trial for either of those getting started in the first half of next year. Data readout from the dose escalation will be later this year, data readout from the salivary gland expansion late this year, and then breast cancer early next year.
Okay. All right, then. We've got some dates for your diaries then. Let's move on, shall we? Because we have some questions on the AVA6103.
We've got just a few questions on this one. Looking at the exposure, it seems that the blood level in the graph still remains higher than the desired concentration, the IC50. Doesn't this worry about toxicity? Did the animals have any toxic effects?
Good question. One of the things that I'll start with, and I'm going to pass this over to Michelle. The IC50 is actually a concentration that shows where the drug is able to kill, what concentration can it kill tumor cells in a dish. It actually has nothing to do with toxicity. There we have to go back, way back in history, in the early 2000s, and look at those initial trials with exatecan in the clinic. What those trials tell us is that the toxic level that we see in the bloodstream is actually significantly higher than that IC50.
We are not even reaching those levels. Michelle, maybe you could give a little bit more color around what we have seen with the animals and how the dosing has been compared with exatecan alone.
Yeah, no, absolutely, Chris. I think that is a really important point about the clinical trials with exatecan, that there was some therapeutic window there. They did see some low levels of efficacy at doses that were tolerated. What we are doing is really trying to expand that window. In all of these animal models we are presenting, the animals are not experiencing any kind of severe side effects. The animals are healthy. In fact, the data we have just presented was a relatively low dose of 6103, so very well tolerated. The data we just talked through was also a low FAP model.
We have seen much bigger differences in some of our models that express higher levels of FAP, and that was presented on our AACR poster, up to 85 times different in terms of the level. I think also important to look at the shape of the curves. What we're seeing there with that exatecan as well is that real extension of the exposure. That was one of the reasons it did not work in the old trials with exatecan, conventional exatecan, was the half-life was very short. What we're seeing is we are extending that half-life. I think that's the really important point, is that the length of time the exatecan has to kill the tumor cells will be much longer. Yeah, all our experience is that this molecule can be dosed at 6103, can be dosed at a much higher level than conventional exatecan.
Very supportive of this sort of favorable safety profile.
Okay, let's move on. We're obviously sticking with 6103 for a second. What indication will the company pursue registration for, and how will you make that decision?
Registration is a little bit further down the pike. What we are focusing on right now is really running a very efficient phase one trial and really limiting the disease setting so that we are walking in and putting the drug right into patients that we would expect to respond. Michelle has really led this with the translational group. Michelle, why don't you talk a little bit about how we're picking that small set of diseases to go forward with?
Yeah, thanks, Chris. Maybe we should go back and talk about the Tempus collaboration again. Very excited about this.
One of the ways we're doing that, as we discussed earlier, was looking for patient populations using this very large database from Tempus and the AI capabilities to identify the patients who express high levels of FAP, but also express markers that indicate that these tumor types will be sensitive to the mechanism of action of exatecan. What's really interesting is that some of the tumor types we've identified, small cell lung cancer, pancreatic, gastric, and cervical cancer, which we talked about on our AACR poster, all of these also are places TOPO1 inhibitors are used. We have a really strong rationale to move forward into these tumor types that are sensitive TOPO1 and also pretty much across the board express high levels of FAP.
What I think is also really interesting when we look at our preclinical data, we're using some of these patient-derived xenograft models. These are models that essentially they're biopsies and samples from human patients that are then grown in mice and successively passaged in mice. Then we test our drugs. They're like little avatars of humans. What we've shown in some of those models is that we get complete regressions of the tumor with 6103. These are models that don't respond very well to the less TOPO1 inhibitors. We have a real reason to believe that we can improve on some of those existing treatments in these tumor types. I think it's a combination of the clinical strategy, the translational work, and also the preclinical models that together make that package that guide us towards those indications in the clinical trial.
Right.
I love when you explain it that way. The breast cancer phase two with exatecan, you'll recall we've talked about this one. They showed responses even in patients that had been treated with TOPO1 inhibitors. It's one of the trials that really sent us down this path just because exatecan is so potent. It can overcome even resistance to TOPO1. i think that that's—I think our animal data are really, really important there. I think that's saying something that's really important about this drug. I'm so excited to get this one into the clinic. I know you are as well.
Okay, this person here, it's the last one, actually. They want more detail about what is required for AVA6103's preclinical results to be replicated in human. For example, what are the challenges the new capping group and linker may see in human, if any?
It's always a challenge taking preclinical data and trying to translate it essentially into the clinic. One thing that I will say, though, before I turn it over to the card-carrying scientist here, is that it's pretty rare to say the following in oncology research. With AVA6000, the drug worked better in patients than it did in the animal models. It's one of the challenges that we have in terms of our animal models. If 6103, if we can say the same thing about 6103, this drug is going to be a beast. Anyway, Michelle, I'll let you—we'll let the card-carrying scientist answer that one.
Yeah, thanks, Chris. I think in drug development, there's always unknowns. When you have a novel molecule, by definition, there's things that you don't know about it.
What we are very fortunate with 6103 is there's a number of things we do know. We do know how the precision linker behaves in patients because we've seen that with AVA6000. We do know a lot about exatecan, the payload. Again, we've got data from hundreds of patients that's published that we can really understand some of the PK, PD aspects of that. To really understand some of the more novel aspects, we do a really very thorough program of work. Some of that we've presented at AACR, but there's also the non-clinical safety and toxicology package where we'll be characterizing the PK of the molecule in more detail before we go into humans. That will all be done in discussion with the regulatory agencies to allow us to move forward into clinical trials.
We'll have done quite a lot of detailed characterization. I like to think of this as a picture of a jigsaw. There's always a couple of pieces missing, but we'll get to the place where we can see what the picture is before we move into patients to really do with that as much as possible. There's always unknowns, but we know a lot of things about this platform and about the payload. That's very helpful in this situation.
Excellent. Thanks.
The completed puzzle is on the way. Thank you both very much, Chris and Michelle. Much appreciated. That was fascinating and a real education. Thank you as well for sending in all of these questions. They were really, really great from your investors, from your followers, and those people that are watching what you're up to as a company.
Thank you all.
Yeah, thanks for having us.