Super excited to have CytomX Management join us. I'm very intrigued by your colorectal data sets, but I'll let you kick things off and tell us about the construct.
Yeah, thanks, Umer. Thanks for having us here today. So Sean McCarthy, CEO and Chairman of CytomX. And yeah, we're super excited about the year that we've had so far. Just by way of background, CytomX is a. I refer to us as the original masking company, a company that has pioneered this concept of protease-based masking of therapeutic antibodies and other biologic modalities. And this year has been very much characterized by our first look at phase one dose escalation data for CX2051, which is an antibody-drug conjugate targeting EpCAM, a very highly expressed tumor antigen that's particularly highly expressed in colorectal cancer. This ADC is a masked ADC that we have conjugated to a topoisomerase-1 inhibitor payload. It was carefully and specifically designed to be active in late-line colorectal. And EpCAM, to those of us in the oncology space, it's kind of an old friend.
It's a target that's been around since 1979. It's been very difficult to drug because it's present on normal tissues, and we're the first organization to show the significant reductions of tumor volume in CRC patients across multiple doses, and the first to achieve systemically active levels of an anti-EpCAM therapeutic ever, so we're really excited about where we are and the future prospects for what this drug can do for CRC patients and potentially other tumors.
Outstanding. So I've been tracking the CytomX story over time. I remember there was a CTLA-4, masked CTLA-4, which theoretically made a lot of sense. I think there were always some questions around where the unmasking happens. Does it happen on the tumor side or not? But putting all that aside, I feel like with this specific construct, it does look like however you're delivering it, it's getting to the right place in a sufficient volume. And you have responses now. I think five out of 18 is the number at present. Can you remind us, what's the total number of patients dosed to date?
Yeah, so let me just recap the first data set. So this was our May disclosure, which, as I mentioned, was the first look at the phase I dose escalation. We presented on 25 patients at that time, 23 safety evaluable at the three doses that we're subsequently expanding at, 7.2, 8.6, and 10 mg/kg given every three weeks. We were super encouraged in that initial data set in the first 18 efficacy evaluable patients to see a 28% confirmed response rate. And this is in a patient population in late-stage CRC with a median number of prior therapies of four. So it's actually a fifth-line patient population. So a really strong start. We have subsequently expanded. We announced in August that we had achieved our initial goal of expanding each of these three dose levels to about 20 patients per dose.
Then on our earnings call a few weeks ago, we further updated that we're taking enrollment in the phase 1 study to 100 patients. Our next update is planned for Q1 of 2026.
Realistically speaking, how many of those 100 should be evaluable? We know, for example, seven out of 25 were not evaluable. So of the 100, how many will be evaluable?
Of the 73, we'll have pretty good follow-up on those 73 patients that we mentioned in August. The total evaluable is going to be somewhere between 73 and 100. We're actively enrolling right now, and so hard to say exactly, but it's going to be a meaningful update.
Got it. And dose levels? So previously, just so I have it right, 2.4-4.8, there were only two patients, so one at each dose. Is that right?
Correct. They were single-patient cohorts, just to speed through escalation.
I see. So you went straight to, and there has never been a DLT or anything of that sort.
We didn't see any DLTs in dose escalation at that stage. And in fact, we had escalated beyond the five dose levels that we reported in that first disclosure to dose level six and seven. Those dose levels we sensed at the time when we reported were probably reaching maximum administered dose. And so we currently are, and we continue to expand those levels three, four, and five of 7.2, 8.6, and 10.
You went higher than 10.
We did.
Which is where that DLT was seen. Is that right?
No, again, we didn't see any DLTs as of that data cutoff of April that we reported in May of this year.
Okay. I'm hearing there might have been a DLT since then, or am I reading too much into that?
You may be reading too much into that.
Okay, got it. So I know you mentioned dose levels four, five, six, seven, but just very simplistically, 7.2, 8.6, and 10, those are the three doses, correct?
Correct. Those are the doses that we're currently expanding, and our goal is to select one, potentially two of those doses to move forward into the next phase of development next year.
Excellent. From your perspective, what type of response rate would be sufficient for forming the basis of go forward for one dose over another, or are you balancing that versus the safety profile?
If you look at the current treatment options for third-line or later colorectal cancer, they're pretty dismal. We know that response rates are in the low single-digit arena. And we also know that progression-free survival is in the two to three-month range. That's a little bit better for the combination of Bev and Lonsurf in the third line, where PFS is about five and a half months. But again, back to the data that we showed, 28% confirmed ORR, 94% disease control, 5.8 months of PFS. That gives us a lot to work with as we define our go forward strategy for this drug candidate.
The expansions across these three dose levels are intended to allow us to zero in on which of these doses will have the optimal profile for moving into what we anticipate the next study to be a potentially registrational study.
Got it. Is there an expectation among your investors that the response rates stay in the 20s as you put out the next update?
I think expectations around ORR are always high. Again, I think we have delivered something we think very meaningful with our first look at this data. When we look at the benchmarks, we see a clear path to moving forward into a potentially registrational study with a high probability of technical success. We're not at this point guiding to any specific numbers.
Got it. Sean, do you have visibility on data cutoffs beyond May?
We do see. We are tracking at a high level ongoing data. This is an open-label study. It's obviously very much work in progress.
Got it. As it relates to PFS, is there a certain number out there? Is there a certain number you want to see?
Again, it really depends on where we go in terms of line of therapy. In a way, the most obvious next step for the program would be a study in the fourth line, where we would take two doses of CX-2051 randomized against a control like fruquintinib, which is approved in the fourth line as an ORR of 1%-2% and PFS of about three and a half months. So you can see there's a wide margin, a very wide margin on both response rate and PFS that we have for our drug so far. We think that's a very compelling study design. That's one thing that we're evaluating. We're not, however, ruling out the possibility that the next study could be in the third line, monotherapy against Bev/Lonsurf or potentially monotherapy Bev versus Bev/Lonsurf.
And for that reason and others, we announced on earnings a few weeks ago that we have a plan to initiate the Bev combination phase Ib study for the combination in Q1 of 2026.
Got it. If you go down that fourth line study, you could have it wrapped up in a year or so, which could even be a potentially registrational trial versus Fruquintinib. Isn't that a reasonable statement?
We certainly think that study could be executed very efficiently, given how rapidly patients on the control arm would be expected to unfortunately progress.
Got it. But it now also sounds like a third line study as well is under planning or instead of a fourth line study?
I don't know that it would necessarily be as well. I think we just have to, right now, we're keeping all of our options open, and we'll be making data-driven decisions as we move into 2026.
Got it. As it relates to additional indications, what do you have in mind?
One of the great things about EpCAM, in addition to being such a highly and consistently expressed antigen in colorectal cancer, is that it's present on most solid tumors. So there's a multitude of options for the next study. We are highly motivated to begin work in non-CRC tumor types next year. Examples include other GI tumors like gastric and pancreatic. The target's highly expressed in non-small cell lung cancer, in endometrial cancer, and many others. Many different options that we're currently triaging. No decisions made yet, but we do plan to initiate at least some of that work in 2026.
Excellent. Excellent. Excellent. Maybe just backing up just a little bit then, I want to understand, is it topo doing the magic, or is it masking doing the magic, or is it a combination of both? But maybe just let me just go step by step, starting with the payload. This is not just the same topo payload that's been used in the other ADCs. So could you remind us what the potency differences are?
Yeah, so this is a novel payload we licensed from ImmunoGen, a novel linker and payload. The payload we call CAMP59. So it's a novel CAMP2D derivative that was designed and optimized by ImmunoGen to be similar to the payload on Enhertu, so DXD. And it behaves very similarly in vitro. It has a similar potency, similar anti-tumor activity on cells. And the linker is a cleavable peptide linker. It's a trialanine linker. So it's ALA ALA ALA CAMP59. And that's been designed to optimize bystander effect. We do expect that payload to be—that linker to be cleavable in the tumor microenvironment and in tumor cells. So that said, this is the first in human experience with this linker and payload, and we're learning as we go. And thus far, it's behaved really very well. In terms of the masking, let me just comment on that for a moment.
Again, we're just so excited about what we've seen so far. This really is, we think, a breakthrough in the treatment of late-line CRC. We're highly motivated to bring this drug into earlier lines of CRC treatment, not to mention other tumors, as we just discussed. And we really believe it's the combination of, it's the right target. EpCAM is certainly a wonderful target for treatment of colorectal cancer if you can drug it. The right technology, which is masking to open a therapeutic window, and also the right payload, which was carefully selected given what we know about the sensitivity of CRC to topo one inhibitors, given that Irinotecan is standard of care in the earlier line setting.
Got it. Are you aware of any competitor data sets which have tracked in ADC space, tracked in double-digit response rates in an advanced line setting?
This is the first anti-EpCAM ADC that has ever been developed. In terms of the target itself, we're a long way ahead.
But on any topo-based ADC?
Yeah, yeah, yeah. So with regards to other mechanisms, there are a couple of other ADCs with topo one payloads beginning to also show promise in CRC. One is ABBV-400. That targets c-Met with a topo one inhibitor. And then that's being developed by AbbVie. And the other is M9140, which is targeting CEACAM5. That's being developed by E. Merck and has also shown promising activity. So I think one of the key developments really for CRC over the last several years is these advances in ADCs. After many years, I think we can say that antibody drug conjugates are coming to colorectal cancer, specifically topo one ADCs. We believe we have a really, really good one, a really differentiated one that goes after what we think could very well be the best target.
Are those response rates as robust, in your opinion?
I'd say they're in the same general ballpark. I think we feel that our initial data set is probably stronger. But there are respectable response rates being put up, I think, by those drugs.
Excellent. Excellent. Excellent. Fantastic. Maybe in the last couple of minutes that we have, I think it might be helpful just to spend some time on your masked interferon alpha. Could you just remind us what the program is, what's the structure and the indication?
Yeah. So in a way, our CX-801 program, which is our masked interferon alpha-2b, it shares some similarities with the EpCAM program in that both of these targets or pathways we've known about for a long time, it's all a question of how do we harness their true potential. Interferon alpha was actually the first immunotherapy to be approved decades ago. It's fallen out of use because it's very poorly tolerated. It has a high rate of systemic adverse events. So what we've tried to do is to clamp down that systemic activity by dually masking interferon alpha with the goal of maintaining its potent ability to modulate the intra-tumoral immune microenvironment.
We presented preliminary data at SITC about a month ago showing in the first five patients treated with paired tumor biopsies, very powerful activation of interferon signaling in the tumor, in tumor cells, in stromal cells, and in multiple cells of the innate and adaptive immune systems. We also demonstrated induction of checkpoint inhibitors or checkpoint proteins PD-1, PD-L1. And the reason that's important is that we think it sets the stage for the actual drug here, which will be the combination of CX-801 with Keytruda. We have access to Keytruda under a supply agreement from Merck. We began that combination dose escalation also this year. And we are on track to have data in late-stage melanoma in 2026 by the end of next year.
And when you say late-stage melanoma, these are IO refractory patients?
Yeah, post-PD1. So the goal is to restore checkpoint sensitivity to melanoma in the post-checkpoint inhibitor setting. And if you look carefully at the data that we showed at SITC, the drug's doing everything that we hoped it would do in terms of activating that inflammatory microenvironment, including driving chemokine recruitment of large numbers of T cells into the tumor microenvironment. So it really sets up the ideal environment for combining with PD1. We also know from work that Merck has published previously that the combination of interferon alpha- 2B and Keytruda is highly potent in melanoma. The problem is that it's just really poorly tolerated, as you might imagine. So the goal, similarly with EpCAM, is to open a therapeutic window for this well-established mechanism.
We think 801 over time could become a new centerpiece of combination immunotherapy across a wide range of different mechanisms.
Excellent. Any questions before we wrap up? I guess maybe one last one for me is when I was looking at those five responses you guys have shown to date, I think three out of five were at the top dose, the 10 mEq dose. So my bias was most likely the dose going forward will be the 10 mEq dose. Again, it's the small sample size, so we might see more data. But on the flip side, I was trying to balance that with the diarrhea rates seen early in the study. So is that the primary consideration for you as well? You're comfortable on the response rates. You want to sort of make sure you find the sweet spot on managing the GI tolerability.
Yeah, we're going to know a lot more about this by Q1 of next year, given the expansions that we're doing. We obviously have been super encouraged by the fact we've seen activity at all three of the expansion doses, at 7.2, 8.6, and 10. It's hard to tell looking at that first data set where the dose response is, but typically ADCs do show dose responses both in terms of safety and efficacy. And we'll look at the totality of the data in Q1 and determine our next steps. But again, a very strong start for this very novel ADC that we think has definitively shown proof of concept for our masking technology platform.
Excellent. Outstanding. So maybe just as we wrap up, remind us again the venue for that data reveal for the 100 patients. What would that be and what's the cutoff?
So keeping our options open at this stage. It'll be sometime in Q1. So please stay tuned.
Okay. Excellent. Thank you again. Thank you so much for joining us.
Thanks for having us. It's been a pleasure.