Rohit, when we get started, can you just close that door? All right, good morning, everyone, and thanks for joining us at the Morgan Stanley Global Healthcare Conference. I'm Michael Ulz, one of the biotech analysts here, and it's my pleasure to introduce René Russo, CEO from Xilio Therapeutics. Just a reminder, the format for today is a fireside chat, but if anyone in the audience would like to ask a question, please just raise your hand, and we'll make sure to address it. But before we get started, I just need to read a quick disclaimer. "For important disclosures, please see the Morgan Stanley Research Disclosure website at www.morganstanley.com/researchdisclosures. If you have any questions, please reach out to your Morgan Stanley sales representative." With that, René, great to see you so early in the morning today.
Good morning.
Thanks for joining us. Maybe to start, for people that aren't familiar with your story, if you can give us a little bit of background on the GPS platform and some of the advantages of that approach.
Thank you. Yeah, so the GPS platform, which is Geographic Precision Medicine, is designed to take very potent, anti-tumor agents, in particular for us, immunotherapies, and preferentially activate them inside the tumor microenvironment, and what that allows us to do is get very low systemic exposure, so you're protecting healthy tissue and keeping adverse events to a minimum, and then getting that maximum activity in the tumor microenvironment, so we think this is an important way to move forward with potent immunotherapies to get higher doses and, combinations, and even, get into certain targets that were otherwise not touchable.
Makes sense. And maybe if you can just give a little bit more detail on specifically how you do that with the platform. And then is the platform able to be used for different agents? And how different are those masking moieties? Are they different for different programs, and how much additional sort of work needs to go in to design those?
Yeah, I, I would say at this point, we feel very confident that this is indeed now a platform. We've applied this approach, this unique protein engineering approach, that is designed specifically for each molecule. So these are not sort of plug-and-play components-
Yep
... but each molecule is designed specifically with a couple components. First, a masking domain, which blocks the activity of that molecule, so when it's circulating in the periphery, it doesn't bind to any target. We have linkers and protease cleavage sites, and those protease cleavage sites are activated by any one of a family of MMPs that are highly active and present in solid tumors. And in the presence of those proteases, that cleavage site opens up, the mask comes off, and that molecule is active. So we're taking advantage of the fact that solid tumors have high expression of proteases, which are required for growth and invasion metastasis.
Is there any risk of some MMPs, you know, outside tumors, and, and what's the level of that and...?
Yeah, so we do have expression of proteases sort of all over the body.
Yep
... but they're under control of regulators that keep it at bay, so that you don't get tissue destruction. There can be certain, like, autoimmune and inflammatory situations where you can get some protease activity. But to date, we've now tested our molecules in over a hundred patients, now across three different programs. We haven't seen any specific issues related to outside of tumor activation, so we have low levels of circulating activity in the periphery. And even in highly proteolytic organs like the spleen, we looked carefully at that as an example and did not see high activation there. So I think we have... You know, it's nuanced to get that right-
Yep
... family of proteases, so it's not too broadly activated, but is activated across a wide range of solid tumors. And I think that the team has done well.
Yep, makes sense. And is it, is it roughly the same set of proteases across the different, different agents, or is it slightly different?
We typically work with now a relatively standard set of proteases that we have learned from collecting over a thousand human solid tumor samples and analyzing them, what we believe are the key ones, and then we build in redundancy into each molecule.
Mm-hmm.
So we are working with a pretty consistent family of MMPs. What's important is in the 3D design, that those protease cleavage sites are presented in a certain way that they can interact with those proteases. So a lot goes into the structure, and the 3D conformation of the biology is really important.
Yep, makes sense, and, maybe that's a good start on the background here, but maybe we can move into the, your first program, XTX101. That's your masked CTLA-4. Maybe you can talk about some of the key features of that molecule and maybe how it's different from Ipilimumab.
Yeah, so we consider XTX101 to be sort of that one of the next generation CTLA-4s that we are starting to see as a class are having activity in colder tumor types, which have not historically responded to previous CTLA-4s, like ipilimumab. From a molecule perspective, we're starting with a base molecule that has ten times the potency of ipi, both in vitro and in vivo. We have put in several mutations into the Fc for Fc enhancement to drive ADCC, and ultimately, we've now demonstrated that that actually leads to Treg depletion, which was theoretical. Now we can-
Mm-hmm
... measure that. And then, of course, we also have the masking technology, which makes it tumor selective and reduces the amount that's circulating in the peripheral blood. So when we give an XTX101 dose, we typically see about 15% activated in the peripheral circulation and about 70 to 90% activated in the tumor, clinically.
... Gotcha. And you're currently in a phase II study in colorectal cancer, but maybe before we dig into some of that, can you share what you've learned sort of in the phase I and the rationale for specifically going to that tumor type?
Mm-hmm. Yeah, so in the phase I, and this was the first molecule we put into the clinic with a platform, so it was important that we did demonstrate that consistent PK, predictability. We did demonstrate that tumor selective activation in actual patient tumors, measured that, and was able to report that out, so that getting that 70-90% active in the tumor, 15% or less in periphery, was an important learning. In addition, in the phase I monotherapy study, we saw something pretty unique. We saw, a deep and durable monotherapy response in a patient with Stage IV, PD-L1 negative non-small cell lung cancer. And there was a couple of things interesting about this that led us down the road of CRC. First, it was PD-L1 negative, so a colder tumor type.
Typically does not respond to even historical IO combinations, including ipi-nivo. So that was interesting, but this patient in particular was presenting with what was defined as innumerable liver mets at baseline. So these liver mets are sort of a kind of a key problem for MSS CRC in particular, and have historically been very resistant to IO treatment. And in this patient, those liver mets cleared after the first cycle and remained cleared with the monotherapy in a durable manner, and that was pretty exciting to us, and I think also why we ended up deciding to move into colorectal cancer, and why Roche joined us as a partner on that program. And it turns out liver mets are pretty similar, even across different tumor types, biologically, so there's rationale to believe that that should translate.
We also know that liver mets have a lot of MMP activity-
Mm.
which is interesting and may be playing a role here.
Yeah. Interesting. Yeah, so you had a dramatic response in colorectal cancer, MSS patient. You mentioned a few times about the activity in the tumor versus the periphery. Is there sort of an optimal ratio there of tumor activity versus periphery, or is it just, you know, more tumor activity is always better than any kind of periphery?
It is pretty, I think it's target and molecule specific for sure.
Yeah.
For the anti-CTLA-4, we believed, in talking to experts in the field, that it was important to have some peripheral activity and make sure we hit draining lymph nodes in particular.
Mm.
So the masking was designed to have some level of peripheral activity, and it is pretty consistent now across quite a lot of patients, around 15%, and it stays pretty stable. Some of that may be activation in periphery, some of it may be the molecule activating in the tumor and then circulating with a long half-life, as an antibody does. In the case of our cytokines, we'd really want almost none circulating in the periphery-
Mm-hmm, mm-hmm.
especially with IL-12.
Yeah, yeah.
So those are designed differently. So it really is target specific.
Got it. Makes sense. Can you just talk about how you selected your, your dose, for the phase II and, and kind of the exposure relative to ipilimumab that you're seeing in the tumor?
Yeah. So for, when we went through the monotherapy dose escalation, our target dose was originally somewhere around sixty, seventy mg per kg, which we believed would be about equivalent to a 10 mg per kg dose of ipi, if it was fully activated. We were able to get up much higher to a dose of 150, Q6, in a monotherapy study, and that's where we saw that monotherapy activity. In combination with the atezolizumab for the phase II, we did study two dose levels, 75 and 150, and we decided to move forward in phase II with a dose of 100, so in the middle, to ensure that we get the profile that allows patients to stay on long term, because that I think that's been a big problem with other next-gen CTLA-4s.
They're very potent, show some really exciting activity, but the tox is often a problem, and then people can't stay on. So we're really trying to make sure we have a profile that allows people to stay on for multiple cycles, continue the full course.
Yep. And you recently started the phase II study, which you mentioned with, in combo with the atezo. Maybe just talk a little bit about the design there.
So yeah, this is right now a phase II, what we call a proof of concept study. We are enrolling patients with late-stage MSS CRC, both patients with and without liver mets, which likely have different response rates based on sort of historical data. So we're including both, and we will look at a total of 40 patients in this initial POC before we then go on into expanding that into a full phase II.
And you mentioned this is in partnership with Roche. Maybe just comment on the partnership and their level of involvement.
Yeah, so this is a co-funded trial, and so each of us are providing our drug. We're sharing the cost of the study, but we are the sponsor running the trial.
Gotcha. And plan to share some initial data in the fourth quarter of this year, so maybe just give us a sense of what we should expect and kind of what you'll be looking for in that early data set.
Yeah. So what we expect to have in Q4 of this year is a readout on the first twenty patients in phase II. Again, it'll be both with and without liver mets, and we expect that it will be likely through the first scan for most of these patients, so it'll be an early read. But we should be looking at the first twenty patients, both with and without liver mets.
And just maybe talk a little bit about safety and kind of what you've seen in the past, and then what's common sort of side effects with CTLA-4s?
... Yeah, so the biggest challenge is, has been colitis for these molecules, and it really can cause patients to come off therapy. So this is what we've been really paying close attention to. We found in the phase I monotherapy that the one fifty Q six had very little colitis, and patients were able to stay on for a long time, and that was key. And was highly differentiated when you looked at that versus, you know, similar doses-
Mm-hmm
of other molecules. So that was key. We also saw with the masking technology, things like very low levels of skin dermatitis, which often happens very often and frequently with CTLA-4. We saw no endocrine AEs, which also are pretty common, so a very differentiated profile. But I think the colitis is the main thing that causes patients to come off CTLA-4 combos with PD-1s, and that's what we want to keep an eye on and make sure stays low.
Got it. You also have guided to presenting another sort of update from this study in the first quarter of next year. Maybe just, you know, talk a little bit about why and sort of what additional data we expect to get there and how to think about that.
So in the first quarter of next year, we expect to have the full 40 patients from that POC, including at least second scans on the first 20. So we'll have, you know, a little bit of longer-term follow-up on the initial 20 and have that full 40 available in Q1.
Can you talk about the bar in terms of response rates in MSS colorectal, and kind of what would be a good response rate?
I think we would categorize the patients into those with and without liver mets and think about them a little differently.
Mm-hmm.
In patients without liver mets, we would hope to see something in the range of a 15% to 20% response rate. In patients with liver mets, historically, you know, the data's been 0%.
Yeah.
So I think if we see any signs of response, that's going to really matter in that population. And potentially over time, things like PFS might matter in that population, but we'll be looking for any signs of activity in patients with liver mets. And that is really, I think, where the highest unmet need remains.
Yep. And assuming you see positive results from your phase II, like, how do you think about next steps and, you know, when could you move that program forward?
Yeah. So if we like, the results from the first forty and we look like we're seeing activity, we would then expand that into a full-size phase II trial, which would allow us to get better sense of overall response rates and really then better design a pivotal trial based on that.
Gotcha. Maybe just talk about the market opportunity in MSS, CRC, and kind of the unmet need there.
Yeah. Unfortunately, this for some reason, we don't really understand why, this cancer type seems to be growing very rapidly. It's the second leading cause of cancer-related death overall, and each year in the U.S., about 90,000 patients are diagnosed with Stage IV MSS CRC, which is a problem because in Stage IV, you know, the choices become very limited, and this is where we're focusing initially. The other issue is that we're seeing patients diagnosed much younger, so in their thirties, their forties, and getting diagnosed with Stage IV because they're not getting screened typically.
Mm-hmm.
So this is something that we're watching closely, but I think a very large unmet need, and especially in younger patients, you know, I think they need immunotherapy options that have long-term durability, and we have to really change this paradigm for those patients. That's the goal.
Yep. And if you do see some nice proof of concept in CRC, would you consider maybe advancing in other indications as well, or how do you think about that?
Yeah, absolutely. I think what's key is looking, seeing the combo data, right? So looking at the combo, the safety of the combo, in this case, of the PD-L1. We have seen evidence of antitumor activity in other tumor types-
Mm-hmm
... you know, across the phase I, so, there could be other tumor types that make sense, and lung, in particular, non-small cell lung is on our mind as a potential next step.
Got it
... given the phase I data.
Yep, makes sense. Maybe we can shift to just XTX301 , and maybe that's your IL-12. Maybe talk about what makes IL-12 so attractive and maybe remind us of the history-
Mm-hmm. Yeah.
-there.
Yeah. So some would say IL-12 was sort of the holy grail or maybe is the holy grail for IO, because IO historically has only worked in immunologically warm or hot tumors, and then there's this whole population with colder solid tumors. MSS-CRC is one of them, where we just don't get, you know, any benefit from IO or very little benefit. And so the idea with IL-12 is it has been shown to really be able to turn those immunologically cold tumors hot and then make them sensitive to IO. So it's really designed to open up the field to cold tumors as well and enable combinations. It's historically been demonstrated to have antitumor activity in the clinic, but so incredibly toxic that has led to severe toxicity and even patient deaths in some cases early on.
So, I think everybody's been very cautious, but wanting to pursue this target, certainly.
Maybe talk about, you know, how you overcome those challenges with your platform and then what you've seen sort of in preclinical development that made you sort of advance the program into the clinic.
Yeah. So we learned a lot from masking our IL-2 that we then applied to the IL-12. The IL-12 was a much more complex molecule. The team was able to uniquely sort of design a mask that also released a half-life extension, so short half-life when it's in the tumor. We had a phenomenal preclinical data on this molecule, where we saw complete regressions at exposures that were well below the level of toxicity that we would see in NHPs. So we had a wide therapeutic index, but getting those complete regressions, we knew we were onto something.
Yeah.
Now we are in the phase I, and that is playing out. We recently cleared a dose of 45 micrograms per kilogram. We're now dosing at 60 micrograms per kilogram. We've added a priming dose, which has historically been shown to help with the initial interferon gamma spike that has led to some serious AEs. So we're doing-
Yeah
... a fifteen priming dose, followed by sixty, and this is over a hundred times greater than the prior MTD of-
Yeah
You know, recombinant human IL-12.
Yeah. Can you talk about the priming dose and how that sort of reduces the spike in cytokine?
Yeah. So what's been pretty clear historically, is that when you get the first dose of IL-12 is what really drives that big interferon gamma peak. And then over time, that kind of comes down, but you still have the other PD benefits that you're looking for, in immune activation. And so it's sort of been learned over time that if you give a priming dose, that you can then blunt that initial interferon gamma peak. We did that sort of proactively once we get to sixty-
Mm-hmm
... because it's such a high dose, and wanted to, you know, prevent running into any problems and just make sure we were being, you know, keeping sort of safety in mind at these very high doses. So that's something we did proactively, and we'll have. We expect to have data on all of this, in Q4 of this year, so we'll look at that profile.
Yeah. You mentioned also sort of things are sort of progressing as expected. You keep those escalating. You're currently at sixty. Can you give us a little bit of detail on what you're seeing and, you know, maybe in terms of safety or any other kind of endpoints there?
Yeah. So what we're seeing is, you know, a safety profile that we're very excited about to get to these doses. You know, we were happy to get to 15 initially, then getting to 45 and 60. I think we're very excited about what we're able to do. So that shows us the masking is working. We're seeing very minimal peripheral activation.
Okay
... less than 1%. So we're not seeing, which is good because we have the short half-life IL-12 once it's activated, so that's working. It's not circulating around. We also have been looking at tumor PD, and we opened a PD-specific cohort as well now. So we're, this is the phase I-B, so we will be looking at what's happening inside the tumor, and have that data available as well.
And are you considering going higher than 60, or are you pretty, pretty high there, and that-
Yeah
... that's enough?
It's a pretty high dose. I mean, we'll. I think we'll always follow the data. I can't say we wouldn't go higher, but I think we're at a, you know, a very high dose and certainly well beyond where we thought the beginning of the active range would be, based on the preclinical data.
You mentioned sort of giving the data in 4Q. Is that going to be the full data set through 60 mgs, or just how to like-
Yeah
... maybe give us a sense of what we should expect.
Yeah, I would expect that data set will include the 60 microgram per kilogram dose. I'm not sure if it will include anything beyond that or if we will dose escalate beyond that, but it should include all safety, PK, PD data we have at that point through 60.
Remind us what tumor types are included in that phase I? I assume it's a range.
This is all comer-
Yeah.
Yeah, all-comer solid tumors.
Maybe just talk about next steps. You share the data in Q4, and you kind of see the signal you want to see, you know, other particular indications you're thinking about or, you know, what could the next steps be there?
Yeah. So I think that the next step would be then, moving into phase II, moving into combinations.
Okay.
I think this particular IL-12 looks very positive for combinations with this safety profile. Be it PD-1s, we like the combination of IL-12 and CTLA-4.
Mm-hmm.
In particular, it's been very interesting, so I think we'll be looking at to open up combinations and initiate a phase II. We will be looking at a range of tumor types that we're deciding on actually now, of course, in partnership with Roche, and it includes some colder tumor types, as well as some, you know, typical warm tumor types.
I guess you partnered with Gilead on this program earlier this year. You made the announcement.
Oh, sorry, I said Roche.
Yeah.
Gilead, yeah.
You caught me there-
Thank you
... for a minute as well.
I need more coffee.
So maybe just talk about how that partnership came about, you know, what drove it, and maybe a little bit of the structure of that deal.
Yeah. So, you know, certainly Gilead has deep expertise and is really focusing on building out their oncology pipeline and has some really interesting agents that make sense to combine with IL-12. And I think they are very, I would say, passionate about this target, and they understand it really well.
Mm-hmm.
So it makes sense strategically. It's an option deal, so this is a scenario where, you know, upfront, we had received a total of about $47 million upfront. We are conducting the phase I, and then, the phase II, we will provide an initial data package that includes the phase I and initial phase II data, and then there is an option for $75 million at that point. There are a couple other components. There is a near-term milestone that's pre-option, that's $17.5 million, and another $8.5 million in equity that we can call at, at timing up to us up until March of next year.
... Got it. Maybe we can shift gears to some of your other pipeline activities. Maybe we could move to XTX501. It's your bispecific. Maybe highlight, you know, some of the preclinical findings there and strategy for that program.
Yeah. So I think what we aim to do with the first programs is just see if we could really apply this masking technology-
Mm-hmm
broadly and lock it in and make sure it works. And now we've treated, you know, well over a hundred patients across all of these programs, and we feel like we can do that. And so now the next step is building these multifunctional molecules where you can look for that sort of combination within a molecule, which gives you much more than synergistic activity, because you're bringing both, you know, in our case, the PD-1 and then the IL-2 to directly to the-
Mm-hmm
... right cell-
Yeah
And then keeping it there.
Yep.
And the results are off the charts in terms of, you know, beyond the synergy of the two.
Yep.
So I think this is definitely the future of multifunctionals, with the masking, and the masking is key. So PD-1/IL-2 is our 501 molecule, are exciting. There's quite a few sort of in development. There's been some recent data coming from some of them in-
Mm-hmm
Indications like MSS CRC, but toxicity is a major issue 'cause they're not masked.
Yeah
... and PK, because the IL-2 drives the PK, which-
Yeah
... can be, you know, not very helpful.
Yeah.
I think the masking is key, and this is an agent where, you know, we are excited to move it forward, and I would say it's the beginning of the next generation of our platform, which will also apply to things like cell engagers.
Yep. Before we move to the cell engagers, can you just talk about mechanistically, why combining PD-L1 and IL-2 makes sense?
Yeah, I mean, it's sort of like the classical, first step. I think, I think we believe we could combine IL-2 with that signal, with, many different types of immunotherapies-
Mm-hmm
... whether it be checkpoints, whether it be cell engagers, whether it be TIL, like, it makes sense-
I see
... in a lot of places, but I think the classical starting with PD-1 makes sense, because there's so much data that you can compare to, to see if you're really getting, what are you getting out of the combination? So I think it's just, you know, starting from the basics there.
Yeah.
And bringing the IL-2 to those PD-1 positive cells and then keeping it there-
Yep
... in the tumor while the proteolytic activity unmasks is really key.
Gotcha. Okay. So the view there, you know, sort of IL-2, and then you sort of use other agents to kind of target the different types of cells, or what are the other sort of components there that you could expand to in the future?
Yeah. So with the platform, I would say now becoming a true platform, we can, we have a lot of white space.
Yeah.
So we're looking at IL-2 as a co-stim. We're looking at other agents.
Mm-hmm.
It's really pretty broad at this point, which is exciting.
Yeah
... and with the masking, I think we can really open up the therapeutic index of molecules like cell engagers-
Yep
... on their own-
Yep
... and then in combination with co-stimulatory agents and multifunctionals.
When can we see some data from that platform and kind of really get that proof of concept?
So we expect Q4. Also, that this will be part of our updates, is to kinda unveil a little bit more of what we're doing with masked cell engagers, effector enhanced masked cell engagers, and, put all of that data out, likely in the fourth quarter of this year.
Okay, and maybe just talk a little bit more about sort of the cell engager platform and your approaches there. I think you have two different approaches, if I recall.
Yes, we have two different approaches that we call Attacker and Seeker. One more straightforward, a more simple one that includes the co-stimulatory agents as well, the effector enhanced. So you'll see both versions of those. You know, in some cases, just masking an existing cell engager can dramatically change its profile, and that would be on the Attacker side, and then Seeker being bringing that effector enhancement with masking also into that same molecule.
In terms of the masking and sort of tumor activity versus periphery, like, how do you think about that for these types of programs?
Yeah, I would say it's, again, very specific to each target. In some cases, if the target is widely expressed-
Yeah
... you know, all over the body, we're gonna be wanting to lock that down. In some cases, if there is already some tumor selectivity, you could probably be a little less cautious about it.
Mm-hmm.
But I think it will be very much based on that specific target, and what we're trying to accomplish and what is the indication.
Yep, makes sense. Maybe you can just talk about, you know, current cash today and your runway and kind of opportunities to sort of extend that, and where does that get you through all these sort of upcoming, you know, proof of concept readouts you have?
Yeah. So the current cash runway, so as of June thirtieth, we had $75 million in cash. That takes us into the second quarter of next year, so beyond the full 40-patient POC readout for the first CTLA-4 program. That doesn't include the $17.5 million near-term milestone on the Gilead deal. That is pre-option, and it doesn't include an additional $8.5 million in equity from that deal that we can pull down. So there's an opportunity to extend with those two components as well.
Gotcha. And maybe just in the last few minutes we have here, if you can just sort of, you have a lot of data readouts coming up, maybe just recap for us kind of the progression from here into, you know, mid-next year.
Yeah, so in the fourth quarter of this year, we should be busy.
Yeah.
We will be sharing the CTLA-4 phase II data with atezolizumab and MSS CRC in the first 20 patients from that study. We will share the phase I data from the IL-12 program, likely through the 60 micrograms dose, so safety, PK/PD, and hopefully some tumor biopsies. And we will be also unveiling what we're doing with the masking as applied to cell engagers, both in this Attacker and Seeker format. And then in the first quarter of next year, the full 40-patient data set from the CRC proof of concept study with atezo.
Okay.
That should have some later follow-up and at least second scans on the first twenty patients.
Okay, great. So we're looking forward to seeing all the data, and then looks like you'll be busy the next couple quarters. So, thanks again, René, for sharing your time with us today. We really appreciate it.
Thank you. Thanks for having us early in the morning.