Hello and welcome to day three of Citi's Oncology Leadership Conference. Delighted to have you with us again. Equally delighted to introduce our first speaker, René Russo, who's the CEO of Xilio. We're a broad church here at Citi's Oncology Conference. We've been hosting companies as large as Genentech, Roche, Sanofi, GSK, now we're down into the smaller names. We think it's an interesting story and a very interesting technology, we wanted to be early to highlight it to you. Unlike the other sessions where many of you are very familiar with the underlying technology, what we've done is allocate some time for René, who's the CEO of Xilio, to really provide some background on the company, then we'll have time for Q&A afterwards, which I'll moderate.
As always, if you do have a question you'd like me to ask, René, then please email me Andrew.Baum@Citi.com, and I'm more than happy to ask on your behalf. With that, René, over to you, and thank you for joining us today.
Thank you, Andrew. Hello, everyone. I'm going to start by sharing a few high-level slides just to orient everyone to Xilio's technology and where we are in terms of our pipeline and what's to come this year. Okay. Hopefully, Andrew, you can see these slides.
Yeah, that's worked just fine.
Perfect. Xilio Therapeutics is a company that is using tumor selective activation technology to take what's different and unique about the tumor microenvironment and utilize that difference to activate very potent IO therapeutics locally in the tumor with the goal of improving the concentrations, the doses, the activity, and the tolerability of highly potent IO therapies for patients. We have 3 clinical stage programs, and I'll highlight to you a few of the upcoming milestones that we have planned for 2023. With our 3 clinical programs, we're anticipating 3 different clinical data readouts this year. The first is for our anti-CTLA-4 program, where we anticipate sharing the full data package from the phase 1a activation trial. That will be in Q2 of 2023.
The second milestone will be in Q3 of this year for our tumor-activated IL-2 program, where we will share the preliminary efficacy and safety data from combined phase 1 and some of the patients from the phase 2 trial. Finally, we will share in Q4 of this year the safety data, preliminary safety data from what we anticipate to be the first three dose levels in our tumor selective IL-12 program. To briefly orient you to the technology, I'm showing you here one of the examples of our molecules. This is a cytokine that you see here. This is an IL-2 example. The way we activate and mask these molecules and activate them in the tumor microenvironment is through a number of design components. First, we take the effector domain, which is a potent cytokine, and depending on that cytokine, we may further engineer it.
In the case of IL-2, for example, we use a beta/gamma approach. We then put a masking domain on that effector domain you see here in orange, that blocks that effector domain, the active cytokine in this case, from binding to any binding site and renders it completely inactive while it's circulating in the periphery. This is all held together by a protease cleavage site you see here in green. That protease cleavage site is the switch that is activated inside the tumor microenvironment by tumor specific matrix metalloproteases that are highly active in tumors. That cleavage site is then broken open and releases the active cytokine locally in the tumor microenvironment, allowing us to get to very high doses and safely administering those high concentrations inside the tumor.
Our pipeline is all in Phase 1 at the moment, so our anti-CTLA-4 program is through the final dose level in the Phase 1a. Our IL-2 program is also nearing the completion of the Phase 1, and we'll be opening up the Phase 2 trial in the Q2 of this year. Finally, for our IL-12, we are initiating the Phase 1 and expect to be enrolling in that shortly. Behind this, we have a number of other programs which we call multifunctional molecules that take both the antibody and cytokine components that we have developed and put them together into a single molecule that has multiple functionalities and is also tumor selective. Let me orient you quickly to our anti-CTLA-4 molecule, which was the first we put into the clinic.
This is an Fc-enhanced, highly potent anti-CTLA-4 that is also tumor selective. By working on that Fc-enhancement to drive additional effector function, the theory is that this will help to deplete Tregs in the tumor microenvironment and be an important combination therapy. We can demonstrate here just quickly in vivo, you see on the left that a very low dose of our anti-CTLA-4 molecule, as low as 0.3 milligrams per kilogram, can generate complete responses. If you compare that to ipilimumab on the right-hand side, that same level of complete response requires 3 milligrams per kilogram. It's about a 10-fold increase in potency in vivo with our molecule. The study design that we are following is a Phase 1a/1b that also can allow for Part 1c.
The Phase 1a is the monotherapy dose escalation in advanced solid tumors. We're currently at a dose level of 150 milligrams every six weeks, and we have opened up the Part 1b expansion cohort where we are collecting mandatory biopsies for PD. It is the Part 1a and some of the Part 1b that will be presented externally in the H2. To date, in the clinical program, we've demonstrated, I think, a number of important things with this molecule. We were able to get to a high dose, 150 milligrams Q6W. We did demonstrate preliminary PK that was dose proportional, so the drug exposures are behaving as we expected. I think importantly, we demonstrated that the masking is working in the periphery, and we're not seeing active drug in the periphery.
We've presented 6-16% active molecule circulating in periphery, it's remaining highly masked. Now I'll turn to our IL-2 program, which is a tumor-selective IL-2. Here is the design of the molecule that we mentioned earlier. The MMP activation in the tumor takes it from a masked state on the left to an active state, which you see here on the right with the mask removed. It has a half-life extension domain that increases the natural half-life while it's in the tumor in the active state. What we demonstrated pre-clinically, which gave us confidence putting this molecule into the clinic, was a number of things. On the left-hand side, you'll see in vivo work where we demonstrated similar antitumor activity with our XTX202, that's our masked molecule, to Aldesleukin, which you see here with, in gray.
Importantly, we were able to get that same level of tumor growth inhibition with the masked molecule as Aldesleukin. However, if you move to the right-hand side panel, we tested the tolerability in an NHP model, and you can see here with XTX202 in red. With that high dose, we were able to maintain no signs of vascular leak syndrome in these NHPs. In particular, we're looking at decreases in albumin, which is the clinical early sign of any vascular leak syndrome. On the contrary, you see the non-masked control, which is here in orange. It is the IL-2 molecule, our beta/gamma IL-2, without the mask. Without the mask, you see very quickly decreases in albumin and that those signs of VLS.
The mask allows us to get the tumor growth inhibition we need on the left and also with a safety profile similar to vehicle and NHPs. This is a view of the study design that we're undertaking for the IL-2 molecule. We're currently in the phase 1a dose escalation in advanced solid tumors. We have achieved a dose level of one milligram per kilogram, which is approximately equal in a ratio of 1 to 1 to Aldesleukin based on in vivo animal studies. We're also looking at now opening up the phase 1b, which is the monotherapy PD cohort, where we will be taking mandatory biopsies, and then we anticipate opening up that phase 2 in RCC and melanoma, specifically, this quarter. Where we've accomplished so far with the IL-2 program is on a number of fronts.
We've accomplished that we've reached the target dose range of 1 milligram per kilogram and continuing dose escalation. This 1 mg/kg was our target dose range because it is similar to the 0.5-1 milligram per kilogram dose level where Aldesleukin demonstrated long-term durable responses. We think high-dose IL-2 is critical. Importantly, and I think differentiated, we're administering this high dose as a completely outpatient regimen with a once every 3-week schedule. The PK is allowing us to dose every 3 weeks, and there is no additional fluid support or any other measures required to administer this high dose in the outpatient regimen.
We're seeing no signs of vascular leak syndrome observed, at, up to this dose so far, and we're seeing tumor-selective increases in CD8 effector T-cells, and I'll show you some of the early data from the first 2 patient biopsies we received. This is the first demonstration of tumor-selective cytokine activation that we're aware of, so we found it important to share these first 2 biopsies as early evidence of platform validation. The first patient that you see here was a 51-year-old male with melanoma. Patient was treated at a much lower dose than we're currently at. This was a dose of 0.38 milligrams per kilogram every 3 weeks. It was in the Phase 1a of the trial. This patient received significant pretreatment, including multiple prior IO agents.
At this dose level, they demonstrated no signs or symptoms of vascular leak, and we obtained this biopsy of both pretreatment and then on treatment, immediately prior to cycle 3. What you'll see here on the left were the first evidence of our platform validation, where you look, first in peripheral blood. The pretreatment specimen is in black, the post-treatment in orange, and you see no change in CD8 effector cells in the periphery on treatment from baseline, which is showing us the masking is working in the periphery.
What we were very encouraged to see on the right-hand side is we did see just over a 3-fold increase in the CD8s and the % CD8s in this tumor biopsy on treatment with no change again in the periphery, so seeing that tumor-selective activation. The second tumor biopsy that we received in this study was from a patient here, number 2. This is at dose level 3. They were receiving a 0.53 megs per K, so slightly higher than the prior patient. This is a 75-year-old male patient with renal cell carcinoma. Similarly, was heavily pretreated, including multiple IO agents. This patient also did not demonstrate any signs or symptoms of vascular leak at this dose. This biopsy was obtained prior to treatment on cycle 2 at day 20, so about 5 weeks after cycle 1.
What you see here on the left is the peripheral blood in black is pretreatment, orange is on treatment. We saw no change in the CD8% in the peripheral blood from the pretreatment to on treatment. Importantly, in the tumor biopsy, we saw a 7-fold increase in CD8 effector cells in this tumor biopsy compared to the baseline. Again, encouraging data to suggest that we are seeing tumor selective activation of our cytokine. These are the two side by side. For the IL-2 program, we're looking forward to the Q3 data readout, where we will be sharing that preliminary antitumor activity and safety data from the phase 1.2 trial. Again, we'll be initiating the phase 2 in Q2 of this year. I'll touch on our third clinical program, which is a tumor-selective IL-12.
We're very excited about the potential for IL-12, both on its own, but importantly in combinations, in particular in combination with IL-2. IL-12 is interesting because of its impact on both innate and adaptive immune responses and the potential to change a cold tumor microenvironment. The cold tumors being typically unresponsive to current checkpoint inhibitors and pushing that into a pro-inflammatory state where you're driving CD8s and NK into the tumor and improving that immune environment to then allow checkpoint inhibitors and other cytokines to work in combination. Today, IL-12s have been incredibly toxic and really unable to be given systemically, but have shown interesting intratumoral activity.
Our approach here is once again to take this molecule and mask it, which you see here on the left in the inactive state with an orange masking domain and the green cleavage site, and then utilize the MMP environment inside the tumor microenvironment to activate the IL-12 locally inside the tumor. I'll show you here some data that's encouraging preclinically. That is, was the background behind bringing this molecule now into the clinic. You see here on the left-hand side that we get significant tumor growth inhibition, even with single doses with our molecule. You see here XTX301. This is a murine surrogate for IL-12, since the mice do not respond to the human IL-12, so we use a murine surrogate.
you see here, even at very low doses, that we get highly, high antitumor activity demonstrated. Importantly, you'll see on the right-hand side that the mXTX301, our masked IL-12, maintains normal body weight, so does not drive toxicity in these animals. However, the unmasked control, which you see here in black, which is an unmasked Fc IL-12, demonstrates toxicity very rapidly. we're able to maintain that antitumor activity on the left without the toxicity seen with the unmasked control on the right. Importantly, we've also demonstrated the ability to have that tumor-specific activation in vivo, which you see here on the left, demonstrating significant active fold changes in the active drug inside the tumor over time.
On the right-hand side, you'll see here we also tested our molecule, which is the mXTX301 in red, to demonstrate that antitumor activity versus what we call a non-activatable control. We remove the green cleavage site. The mask cannot be removed. When that happens, you see you do not get tumor growth inhibition, and it looks the same as vehicle. This demonstrates that it is indeed that MMP activation that is driving the XTX301 activity. I think the most important thing for a molecule like IL-12 is the therapeutic index that we're demonstrating as we decide if this is the right molecule to move into the clinic and decide the doses we should start with.
For this molecule in particular, our tumor-activated IL-12, we demonstrated an exposure range for antitumor activity that you see here in green, and this was done in vivo, both tumor growth inhibition and complete regressions. Then compared that to the tolerable exposure we were able to achieve in NHPs, which you see here in the red line. So we are many, many fold above the exposure range for activity in terms of a tolerable exposure in NHPs. This gives us an estimated therapeutic index if you look at the exposure for safety over the exposure for activity of about 66. As I mentioned, we're initiating the Phase 1a trial for this tumor-selective IL-12. The study is open, and we expect to be enrolling imminently.
The dose we're starting with is a 5 microgram per kilogram dose, which is a high dose compared to prior history with IL-12. We expect to report out data from the first three cohorts in terms of the preliminary safety in the fourth quarter of this year. We will open up phase 1b. We'll be looking at both hot and cold tumors, which is different with the IL-12 program. In the cold tumor cohort, we will be looking to advance this in combination with agents like PD-1 and our own IL-2. So far, to demonstrate with this molecule, we've demonstrated that dose-dependent anti-tumor activity without body weight loss in vivo. We've demonstrated preferential activation in tumors versus plasma.
We have also demonstrated activation in human patient tumor samples versus plasma, and we've initiated the phase 1 trial. Andrew, perhaps I will stop here.
That's, that's great, and appreciate the brevity of the presentation. I know there's lots more data you wanted to share. So maybe just to give a little bit of background, Xilio wasn't just a random company that I suddenly felt inviting to a conference that was populated by much larger entities. What initially drew me to it was a couple of things. Number 1 was the talents and backgrounds of many of your team. So you're from Bristol, Marty Huber's from Merck, and you have a bunch of guys from Novartis. Deep pedigrees in the science with hands that have touched many of the existing drugs. Second, these are clearly validated targets. There's no doubt that IL-2 and CTLA-4 work and IL-12 as well, if you can deal with the toxicity.
There, there's clearly, and the final part that drew me to it was the recent data from Agenus, in patients who have MSS, colorectal cancer, which is, until now, been a relatively non-immunoresponsive tumor, seemingly showing response rates in excess of 40%, depending on patient selection with a non-fucosylated CTLA-4, which is basically half of what you're offering. You're offering a non-fucosylated CTLA-4 with a mask. All these things made me reach out to you and we, through mutual friends, we made contact at JPM. That's a little bit of the context why we had interest. I guess the first thing, before we get into CTLA-4 is to differentiate your approach versus CytomX's approach and the Probody. Obviously, they've had a number of disappointments.
Bristol's recently shut down their masked CTLA-4 trial. They've also killed a couple of the other compounds. What's the difference, biologically between the two approaches? Why do you feel confident that this is gonna yield more favorable data than perhaps they've seen to date?
Yeah. I think CytomX certainly is one of the pioneers in this area of tumor-selective activation. We all have the benefit of learning from some of the work they've done over the years. I think the main difference between our technology and CytomX is that ours is a bit broader and more flexible, and that we can apply our platform technology not only to antibodies, but we see the real value in a tumor-selective activation approach with highly potent agents that have a therapeutic index challenge like cytokines, IL-2 and IL-12. Our technology can also be applied to cytokines and what we call multifunctional molecules that combine antibody cytokine together with a mask. I think that's the number one difference.
In terms of the CTLA-4, specifically, I think what we're encouraged by, that CytomX and BMS seem to be continuing the program that looks most analogous to our program, which is their masked non-fucosylated ipilimumab. That is putting together the components we've discussed that, you know, that Fc-enhanced, and enhanced anti-CTLA-4, next-gen anti-CTLA-4, plus the masking together, we think is likely the right way to go. Where I think CytomX, you know, may have run into trouble, I think is not in the fact that their agents aren't activating the tumor. We believe the data suggests they are getting preferential activation in the tumor.
Perhaps this type of approach is best applied to molecules that really have a therapeutic index challenge and where you know you're leaving efficacy on the table if you could dose higher, if you could get higher concentrations. We think that is an Fc-enhanced anti-CTLA-4, certainly. Potentially not something that is needed initially for PD-L1, which I think is where they initially applied the technology, where you're really not getting additional efficacy from those higher concentrations, and there isn't that therapeutic index challenge that you see with an agent, you know, very different, like an IL-12, right? It's really impossible to give without this technology. I think they have demonstrated tumor-selective activation works. Tumor proteases can be utilized to get that selective activation, and then it's about applying it to the right target.
In terms of where you go from here, and obviously it's early days in terms of generating clinical data, but I presume that you're not gonna partner, and it's unlikely a partner would want to come forward until there is more data in hand at this stage. How are you thinking about building out the clinical development plan, obviously capital allowing? Are you thinking of leveraging the work that Agenus has shown? There's obviously a MSS colorectal is a big indication, and there seems to be a biomarker there in terms of the absence of hepatic metastasis. Therefore, this is an area that you could piggyback off their work, albeit potentially a second to the market. Do you go through the standard basket trial, you know, plus or minus PD-1, and do it like that?
How are you thinking of the gating factors?
Yeah. We are very interested, I think, in the data that we've seen so far from Agenus, and that potential benefit, right, in colder tumors like MSS colorectal with that Fc-enhanced approach to CTLA-4. Whether that's through T-reg depletion or the mechanism, I think we all have theories. I think that combined with recent data from AZ, demonstrating that higher exposure, less frequent anti-CTLA-4 may be the way to go. These are some of sort of the novel views, right, of how this mechanism could be potentially maximized. We, I think want to really be focusing on areas of high unmet need where there's a big impact.
Certainly that approach that we've seen with Agenus has led us to put MSS colorectal right up there on our top list of indications that we would wanna pursue in combination with a PD-1 because of the high unmet need in that area. Absolutely something we're interested in. I would say less interested in following a traditional development path where there's potentially less unmet need.
In terms of how convincing that signal is, and we were talking before the call started that I was surprised when you actually look at the literature, the reported response rates in refractory MSS colorectal for PD-1 is much higher than you might imagine. Therefore, you know, whether this, you know, response rate is a function of selection for prognostic marker, namely the absence of liver mets, rather than any sign of additive or synergistic activity by adding a CTLA-4 to a PD-1. I mean, presumably your team has had the chance to look over the data and whatever there is in the public domain, which is some, but not all. What's your views on whether you think this is just a statistical error and no real different from what you've seen with monotherapy?
Whether you think there's something else going on here?
Yeah. Yeah. Yes, we have been looking closely at this data, and our feeling is that there likely is something additional happening with the Fc-enhanced anti-CTLA-4. We're also curious if that could be further sort of improved with higher doses and masking, really getting that concentration even higher into the tumor microenvironment. I think you're right. We have to be cautious about combination therapy data. For us, our approach is very much focused on first demonstrating clear monotherapy activity with these molecules before moving into large combination trials. I think we want to ensure that we have a highly potent agent, that we're seeing some signs of monotherapy activity.
I think tumor PD will be important to understand what's happening in the tumor microenvironment, and I think getting doses as high as possible. We believe there is something happening, likely, with the addition of the Agenus molecule in those MSS colorectal patients. We do want to further push concentrations with masking, and look for that monotherapy activity prior to going into those combo trials if we can.
You have a collaboration with Merck around your CTLA-4, exploring in combination with Keytruda, and of course, Merck has collaborations with just about everyone on the planet in that context. They also have an equity stake in you from memory, which is less common. There's obviously been some leadership changes at Merck in R&D, and its pendulum has swung a little bit away from the dominance of oncology and particularly IO. What are you seeing, hearing from Merck? They have their own non-fucosylated or fucosylated CTLA-4, which I think they ran as a defense mechanism more than anything else. Could you update us on what you see as their level of interest, whether you think they're waiting for the Bristol data and that will ultimately determine the extent to which they increase their activity with you?
Sure, sure. Well, I can't speak to the broader strategy sort of within Merck, but we do have a long history. They were an investor early on in Xilio and through the IPO. They're fantastic partners in that they have so much knowledge and depth of data that we look to them, you know, to react to the data we're seeing and to help us think through sort of next steps. I think there is interest generally from the Merck oncology team on tumor selective activation, on a potent anti-CTLA-4 and on cytokines, in particular, IL-2 and IL-12. There is broad interest, I think that drove the investment that they've made over the years. Hard for me to say what is their specific interest at this point in any one of our molecules.
Sure. In terms of timing of the last standing Bristol program, with the CTLA-4 masking. Is it from memory the end of this year or is it spilled into next year? Do you happen to know?
I don't know. I don't It hasn't been completely clear to us the timing of when that data might be shared and when that study might be complete. We are certainly looking very closely and anticipating that data. Again, as that molecule probably aligns most closely with what we might expect to see from our own XTX101.
Let's just pivot to the IL-2. There's been a fair bit of pain with some of your predecessors in that space, most recently with Nektar Therapeutics and also Sanofi post the Synthorx transaction. These are obviously differently engineered approaches. Just to confirm, your IL-2 is a non-alpha approach to eliminate the Tregs. Could you just compare and contrast the differences versus Nektar Therapeutics for those who may be less familiar with why that program bombed and obviously it was efficacy, but it wasn't just efficacy. We can talk about other ways of successfully drugging or localizing IL-2 activity. We spoke about the recent Roche transaction with a PD-L1 conjugate IL-2. A little bit of background on IL-2 competitive landscape.
Sure, sure. Happy to. Just taking a step back, I think the reason why IL-2 remains such an important target in the field of IO is that going back to the original Aldesleukin data, again, high dose, and I think this is becoming a theme, right? High dose Aldesleukin, 0.5 to 1 milligram per kilogram. In the patients who achieved complete responses, those responses were long-term durable responses over 10 years. That is what we're aiming to maximize for patients. What we've learned with what I would call second generation IL-2s, I think was important. We saw a number of companies, Nektar in particular, that aims to eliminate the toxicity, right, by giving low dose exposures and really sort of flattening out the PK, in their case through pegylation.
I think with an attempt to just give a little bit of the benefit of IL-2 in combination with checkpoints. What I think we've learned, and was an important lesson is that low dose systemic IL-2, in particular wild type, will drive Tregs primarily. We have to think back to how IL-2 has evolved biologically through evolution in our body. It is not a systemic molecule. It's highly toxic. It is not an endocrine molecule. It's a local paracrine molecule. If it's given high dose in a pulsatile way, I think that brings us back more closely to what we saw with the Aldesleukin responses. We also know intratumoral IL-2 injections in melanoma lesions can lead to a 50% CR rate, which is data I've seen most recently from Memorial Sloan Kettering.
We know that IL-2, if localized into the tumor, can work. I think now we're trying to balance with the learnings we've seen from second generation IL-2s, is getting away from low dose systemic exposure of a wild type molecule, and really finding ways to maximize concentration, localize it in the tumor, and focusing on the beta/gamma approach.
And then-
Yeah, go ahead. Sorry.
No, no. The second part of the question in terms of alternate approaches, and I mentioned the good approach with a PD-1 regulated IL-2. Could you talk to, and maybe describe the competitive or the merits of the different approaches?
Yeah. I think in terms of that local tumor activation approaches, there are masking approaches, like we're taking, right? Where you get the mask is removed inside the tumor microenvironment, and you get that localized effect. The company you're speaking of was Good Therapeutics, where they did a deal recently with Roche on another form of conditional activation. Meaning in certain conditions, the molecule is activated. In their case, they put a PD-1 IL-2 together, which we think makes a lot of sense. That helps to drive the IL-2 to the tumor, stay bound locally, right, via the PD-1, but that the IL-2 was inactive until a molecule binds to the PD-1. Trying to, again, localize and focus that effect in the tumor.
What the Good Therapeutics approach did not have, to our knowledge, is any type of masking. We believe there still will be peripheral activity of that molecule, which may ultimately limit dosing. What we like and what we're looking to with our next generation of multifunctional molecules is similar to that, where you put the antibody and cytokine together, but with the addition of the mask for tumor selectivity. Of course, the other approach is intratumoral, right? Just physical intratumoral injection or administration.
Which has its own limitations depending on indication. Just finishing up on the financials. Just remind us how much net cash you have and what your burn rate is.
As of the last quarter, we had about $140 million cash on hand. We'll be updating this shortly with earnings, and that takes us into the H2 of 2024. It's about a year post many of the milestones that are coming out in 2023. Burn rate is approximately $20 million a quarter.
Yeah. Okay. I guess one thing that's worth noting is, the industry is getting increasingly cold feet about agents which lack single agent activity.
Yes.
I think depending on the outcome of the SKYSCRAPER trial, that may put the death knell in any company taking forward an agent. Without serendipitously, or rather by design, your agents don't fall into that bucket, which is a risk you don't have to worry about. Thank you so much, René. Always a pleasure to see you. I'm sorry it's such an early morning. I appreciate you waking up especially for us. We'll, I'll leave the viewers here. If you do have any questions following the Zoom, you'd like to put to René, please email me and I'm happy to forward it. Thank you again. Operator, could you put René and myself into a side room?