All right. Welcome, everyone, to Jefferies Lending Healthcare Conference, 2025. My name is Roger Song, one of the senior analysts covering MedTech Biotech, in the U.S. It is my pleasure to have the fireside chat with CytomX Therapeutics, and we have the CEO, Sean. Thanks, Roger. Real pleasure to be here. Excellent. Okay, Sean, we have a lot to cover. I think you have exciting—you already reported some very promising data, and then you have exciting 2026 ahead of you. Maybe just give us the state of the union for CytomX. It's been a long road to here, and then, what's the excitement, into the next year? Yeah, happy to, happy to recap, a very exciting year for CytomX. 2025 has been quite exhilarating, actually, as—now we're looking ahead to 2026 that has the potential to be even more exciting, I think, for us. Just to recap who, who we are, what we do. CytomX is a therapeutic antibody company focused on oncology, having really pioneered an approach we call the Probody therapeutic platform. This is a masking strategy that we use to improve the therapeutic window for a range of different antibody modalities in oncology, including antibody-drug conjugates, T-cell engagers, and cytokines. I like to refer to CytomX as the original masking company. We really got this field going, and it's been really exciting over the last few years to see others enter the space and enjoy some level of success as well. We do think that masking is here to stay, based on the work that we've done and others are beginning to do. In terms of this year, we entered the year with a goal, a goal of data on our EpCAM ADC program. This is called CX-2051, and we presented data in May this year, which was the first look at the phase one study for this drug candidate, which we are super excited about, and I think others are as well. We'll talk about some of the details in a moment, I'm sure. We also entered the year with a second clinical program, beginning to make progress, which is a masked version of interferon alpha, and we just presented data at CITSI a couple of weeks ago from the first handful of melanoma patients treated with that drug. It has been a really, really fun year. It's been great to see the overall improvement in capital markets conditions as well, and we've, like others, have benefited from that, and happy to talk about some of the details. Excellent. You know, I've been covering CytomX for a couple of years. You have a very unique technology platform, and you said you're the pioneer of the masking technology. Obviously, we have the field, and I have a couple other followers, but you really pioneered this, this technology. It's not easy to be a pioneer, right? You know, you've been trying different targets, but since recent, earlier this year, you got this EpCAM 2051, this program, giving us very exciting data. Maybe just recap that data for us as well, so we can talk a little about the future expectation for the early next year. Yeah, just to recap our core technology, the basic idea of the masked antibody or the Probody therapeutic is to take an antibody, add a peptide mask to it that is protease cleavable. The whole concept is to engineer antibodies to be protease cleavable antibody prodrugs. We've learned a lot about how to use this technology based on a broad exploration in the clinic across multiple targets over the years. Really, the learnings from that work, and you're right, it's not always easy to be the first or to be the pioneer, but you have to learn, and you learn as you go, and you apply those learnings. With the 2051 program, we've really taken a range of different learnings from the clinic to optimize the design of 2051. Let me just explain what this drug is and what it's shown in our first phase one study. It's a masked 2051, is a masked antibody-drug conjugate targeting EpCAM. EpCAM is a very abundant colorectal cancer antigen. It's expressed at very high levels on pretty much all colorectal cancer patients. That's been known for a long time. EpCAM was actually first described in 1979, partly because it was so abundant. It was relatively easy to find with techniques available at that time. It's been very difficult to drug. Even though there's a lot of it on CRC, it's been difficult to drug because it's also on normal tissues. Masking, our strategy is designed to diminish binding, decrease binding in normal tissues, and allow for binding in tumor tissue because the mask is removed specifically and selectively by tumor-associated proteases. We reasoned that we could perhaps turn EpCAM into a viable drug target for a systemic therapeutic, moreover for an ADC, by taking an antibody to EpCAM, masking it, and conjugating it to a payload, and specifically a payload selected to be active in CRC, which is a topoisomerase I inhibitor. We license it from ImmunoGen, that's called CAMP59, and it's linked to the antibody, to the masked antibody by a trialanine, cleavable peptide linker optimized for bystander effect. Now, one thing that's particularly unique about EpCAM, in addition to its high-level expression, is we know that if you can, you know, locally administered strategies have been effective in treating different kinds of tumors. In fact, there's a drug that's recently been relaunched in Europe called Catumaxomab, which is an anti-EpCAM T-cell engager, which is effective in the treatment of a condition called malignant ascites. This drug has to be given intraperitoneally, but it's really effective in treating EpCAM-positive malignant ascites. We know that if you can get the drug there, it can work. How do you do that with the systemic therapy, when the target's present on so many normal tissues? That's what we've been doing with masking. Our phase one data in the first 20 or so patients presented in May was an incredibly strong start, we believe, in CRC. We demonstrated in 18 efficacy-available patients, a confirmed overall response rate of 28% across three dose levels of 7.2, 8.6, and 10 milligrams per kilogram given every three weeks. That compares to, and this patient population is in the fifth line or later. A very late-stage CRC population where current standard of care results in response rates of 1-2%. In terms of progression-free survival, we had a preliminary estimate of PFS of 5.8 months, which again pretty much doubles the current PFS for standard of care and a disease control rate of 94%. On the safety side, the drug is well tolerated, low rates of hematologic toxicity, and an overall emerging profile that we're learning more about as we move forward. Really excited about that initial data allowed us to conclude a $100 million concurrent financing, with the data released on May 12th this year and put us in a very strong position for the further development of the asset. Excellent. Yes, that's very, very exciting data. Maybe I take a little step back because when you have this promising data, investors, investors starting to ask me, probably asking you as well, is why did that happen, right? You develop, you know, your technology for different targets, different indications, and I think this is the strongest signal you have ever seen. A couple on, on the kind of molecular level. What is the masking technology you apply, the learning from the previous programs you think is different? That's number one. Number two is the EpCAM. How the EpCAM is different from other programs? Because we know a couple other EpCAM programs out there, so they're also in the clinical. Any differentiation there? Lastly, you mentioned this just pretty briefly on the linker and the payload. How differentiated the linker payload you think is well suited for the CRC, maybe other indication as well? Yeah, if I could, it's a great, great series of questions. If I could paraphrase the question, Roger, I think, you know, one might ask what we've been asked a lot this year, which is, why is this one working so well? Exactly. It's a really good question. You know, my answer to the question is, actually, our technologies always work. If you look back at the very first clinical data that we presented, on a drug we called CX-072, an anti-PD-L1 masked antibody, it really taught us a lot about how to mask antibodies. We checked every box that we set out to check in that first clinical trial, including robust anti-tumor activity in multiple tumor types and improved safety in the form of improvements in immune-related adverse events. The underlying masking technology that we're using in 2051 is really quite similar. We're very confident that we have unmasking in the tumor based on a lot of work we've done in the past. I think what's optimized about 2051 is really the combination of the target selection in terms of its abundance and its prior validation. The effector mechanism that we've chosen here to really drive efficacy, which is a topo I payload. Of course, we know that irinotecan is a crucial component of earlier line therapy for CRC. CRC responds to topo I inhibition. And thirdly, we executed a very focused clinical study. We decided, we took a significant amount of risk in this clinical study. We decided to only enroll patients with CRC because we felt like that was the killer experiment. Let's do the killer experiment. That has really paid off. As As opposed to doing a more distributed all-comer phase one study, which may have given us a signal in this tumor or that tumor, we just really committed to the clinical design. I think those three things, the indication, the prior validation of the target, the abundance of the target, and the use of this effector mechanism in conjunction with our previously validated masking technology is really what's allowed us to deliver. Got it. On the antibody side, any difference on the binding, any engineering to the binding or binding site, epitope, and then affinity? We like, we like our antibody, a lot. It has a high affinity for EpCAM and, you know, we haven't said too much about the specific epitope, but we do think it's a great antibody. Okay, great. All right. Let's, kind of, you know, go back to on the clinical side. You got this early data sets and then very critically, you already expend a couple of those cohorts you think potentially can be the therapeutic dose. Tell us about the enrollment so far. Seems you increased the size of the enrollment recently and then the public announcement. What drove that enrollment increase? We talk about the expectation. Yeah, at the time of the data disclosure on May 12th this year, we had already decided to expand three dose levels, the 7.2, 8.6, and 10 megapac doses, based on the really promising activity we saw in those first few patients. Our goal at that time was to increase the N for each of those cohorts to about 20 patients. We announced in August this year that we had achieved that enrollment goal, and we updated on our recent earnings call. In fact, we further increased enrollment from about 70 patients in August to what we anticipate to be about 100 patients by the time we get to our next data disclosure, which is scheduled for Q1 of 2026. We decided to increase the enrollment because for a number of reasons. First of all, there's a lot of demand from our investigators to treat their patients with the drug. This is, this is an active drug. There are very few, very poor options available for late-line CRC patients. A question that we got a lot around the August update was, do you plan to enroll additional patients? We now do, and we are actively enrolling. The other reason, of course, is that as we move towards the Q1 disclosure, what we're mostly focused on, number one, confirming the efficacy signal that we saw in May. Number two, further understanding the safety profile of the drug and integrating number one and number two into number three, which is to pick a dose or doses for the next study, crucial next step, and to go to FDA in 2026 to have a discussion around what the next design that could be potentially registrational could look like. Of course, more data is always good, and we think that having about 100 patient phase one data set is going to put us in a very strong position to determine next steps for this very exciting asset. Okay. Totally agree. Each, a couple of specifics in terms of the patient enrollment, because you know the strong signal you see from the previous data sets, and do you see the difference in terms of the patient, baseline prior therapy? Starts you moving to a little bit earlier line or stay in this fourth plus line? I believe that's the medium, pretty high medium prior treatment there for the first. Yeah, so just to recap, the median number of prior therapies of our patient population enrolled, up to the first data update, is four priors. So it's essentially a fifth line patient population. A few things on patient characteristics, which we think are incredibly encouraging about the 2051 data so far. First of all, EpCAM, we have validated that EpCAM is indeed expressed at high levels by IHC in every patient we've enrolled on the study, at least patients for whom we have an evaluable biopsy. That's consistent with data that's in the public domain. It really is a consistently expressed antigen. You can also see in the public databases, and we validated this in our own work, that EpCAM is expressed across all stages of disease, stage one through stage four, and also whether or not patients have metastases in various organs. It is a very highly expressed, very stable antigen. Moreover, our initial data showed activity regardless of clinical characteristics that are typically used to select patients and drive them down different therapeutic pathways. We have activity in patients that have, whether or not they have liver metastases, whether or not they have KRAS mutations, whether or not they have left side or right side tumors, and in all patients, all patients have high EpCAM. This truly is a drug with what we believe could be pan-CRC potential. That has enormous implications for the commercialization of the drug because the oncologist does not have to do any more workup on their patient before putting them on 2051 in terms of other molecular characteristics. We see that as a key advantage, not just for the potential uptake of the drug, but also for the breadth of the market opportunity. In terms of line of therapy, this is a very big topic of conversation right now with our stakeholders, and internally at CytomX. Of course, we are starting in the late line where the medical need is highest and where we believe the development strategy can move fastest. It makes most sense, we believe, for our first registrational study to be in the late line, fourth, potentially third, and to work progressively to bring this drug into earlier lines of therapy, into third, into second. Our vision, potentially to first line, our vision has always been for 2051 to potentially replace irinotecan in the CRC landscape because, of course, we have a topo I inhibitor ADC. This would be very consistent with what we've begun to see across multiple ADCs, initially beginning in hematologic cancer, but now increasingly in solid tumors where ADCs are coming earlier and earlier in the treatment paradigm and starting to replace chemo by being more effective. If you add that all up, this is a major opportunity across multiple lines of CRC, and perhaps we'll get to non-CRC indications in a moment. Yeah. Yes. And then, you know, understanding you are still enrolling the later line population. And then in that population, current standard of care, the, you know, efficacy basically near zero. So just to remind us, so what's the benchmark from the current therapy given you already show pretty promising, even, you know, better than most people expect, expected efficacy? What do you want to see in this, you know, next year data set for you to have a confidence to move into the, next stage? We've got a lot of room to maneuver in terms of how this next data set comes out. To the benchmark, just to restate, the currently available options for late line, even in the third line, patients with metastatic CRC deliver response rates in the 1%-5% range with progression-free survival of about three months. Patients are progressing. If they haven't progressed on their first scan, they're progressing at their second scan. The options are dismal for late line CRC. We've shown in our initial data set 28% response rate across all three doses. A really wide margin between what we've already demonstrated and what is established standard of care. Again, to reiterate on progression-free survival, 5.8 months was our preliminary assessment compared to about three, three and a half months for current standard of care. That means that a registrational study would not need to be too large, and could be executed very, very quickly, we believe, to get this drug to its first approval. Got it. Okay. And then on the safety side, since you know, we know EpCAM is a target, but the question is you cannot dose that systemically because you picked to say, expressed, but you actually find the therapy window because of the masking technology. One thing stick out is the diarrhea. Tell us about that. And then what's the mitigation strategy you have been implementing to the phase one B? If anything, you can tell us right now how you feel about the diarrhea profile. Yeah. Back to EpCAM as a target and what's limited EpCAM in the past, the classic EpCAM-driven toxicity that thwarted the very first approaches with therapeutic antibodies was acute pancreatitis. In our study, not only have we not seen any pancreatitis signal, but we haven't even seen any modulation of pancreatic enzymes, no changes, in our initial data set in amylase or lipase, suggesting strongly that masking is delivering to open a therapeutic window for, you know, on the EpCAM side of the equation. Now, of course, we have a payload on the ADC, which is a novel camptothecin derivative called CAMP59 licensed from ImmunoGen. And we know that these topo I inhibitors can have a range of AEs. We anticipate toxicity from the payload, including a certain level of hematologic tox and likely a certain level of GI toxicity. On the hematologic side, let's start there. Very attractive safety profile, low rates of grade three anemia and neutropenia, which we think offer a great combinability with chemo in potential earlier line strategies with this drug. In terms of other toxicities, low rates of upper GI nausea and vomiting, which again is a hallmark of topo I inhibitors. That's something that we've seen very little of. We do have a signal, as you rightly point out, we've reported in the first data set, 21% grade three diarrhea. Now, again, topo I payloads well established to cause GI toxicity. We anticipate toxicities. We're not masking the payload. It's really important to emphasize that ultimately we always thought that the MTD, or the maximum assessed dose of 2051, would be limited by the payload. It turns out it's most likely the GI, GI toxicities that we need to pay most attention to. That number of 21% comes from the first 12 months of understanding and evaluation of this drug in patients in whom there was no pretreatment of any kind to try to prevent or slow or prophylax against the onset of diarrhea. One thing that we initiated in Q2 of this year was the use of loperamide prophylactically to try to get ahead of diarrhea in patients because we do want to try to get that rate, first of all, understand the etiology of the, the diarrhea and get that rate as low as we can. That's work in progress. In our update in Q1, in addition to confirming the efficacy signal, which really is the primary goal, and including having an initial assessment or potential dose response, we expect to have additional information on how these prophylactic measures are helping us better understand and mitigate the one toxicity that we need to understand more about, which is grade three diarrhea. Yeah. I know it's absolutely not uncommon for topo I ADCs to have the diarrhea, the tox. You know, understand you're using the loperamide as the prophylactic measure. How standardized that is, you know, how, you know, compensant or standardized, easy for the physician to implement this in their practice and then clinical trial. Yeah, the protocol is relatively straightforward. We haven't disclosed the details. The guidance that we gave our investigators in Q2 was through an updated adverse event management plan. We do give the investigators some level of discretion in terms of how they implement that because, of course, it's up to them to manage their patients. Every patient with late line colorectal that comes onto a study like this, every patient has had their own unique journey in coming onto the study. Again, some level of discretion. I should also say that not only is this not, you know, the first drug to have, some GI toxicity in oncology, by any means, there are other, examples of studies that have been done, actually, you know, controlled studies that have been done to specifically evaluate the, the value and benefit of prophylactic measures in, in ameliorating diarrhea as a side effect. I would point to two studies in particular that I think are relevant. First of all, the prime study that was performed, on Trodelvy, looking at the, rate of grade three diarrhea in patients with, triple negative breast cancer, that was, that was successful. It demonstrated a significant reduction, using prophylactic loperamide in the incidence of grade three diarrhea. The other one is neratinib, the HER2 inhibitor small molecule, where a pretty large study actually that was run, several hundred patient study that was run on that drug, looking at a number of different prophylactic measures, again, in looking in decreasing the incidence of grade three diarrhea, those agents, including loperamide and additional agents, including colestipol and budesonide. There are multiple strategies at our disposal, as we continue to learn more about 2051. I want to just really emphasize that we're really excited, with the data so far in terms of the breadth of the therapeutic window. We've done something that no one's done before, which is actually be able to treat patients with colorectal cancer and achieve objective responses, with an ADC, not only an antibody to EpCAM, but an ADC to EpCAM, which is really achieving remarkable results in a very, very difficult to treat patient population. Got it. We look forward to that data package early next year. Would that package be sufficient for you to talk with the FDA to decide the pivotal path forward or what kind of a pivotal path forward program you are looking for? That's absolutely one of the reasons that we decided to enroll additional patients was to make sure that we have the most robust package from phase one to go and have that conversation. Our goal is to get this drug onto a registrational path as quickly as we can. Awesome. All right. Great. What's the balance sheet? How does that support this program and then the beyond? You can mention a couple, you know, early pipeline, just like you said. Yeah. We just reported cash of a little under $150 million that funds us into Q2 of 2027. So well funded, in a strong cash position. We continue to enjoy, as you know, a number of really great partnerships, including ongoing alliances with Regeneron and with Astellas. We continue, will continue to be active in business development, having recently hired Rachel, who's right here in the front row, Rachael Lester, our new Chief Business Officer, and continue to be open for business on the BD side, on platform and programs. We've always found ways to robustly finance CytomX, and I'm sure that will be no different in the future. Excellent. Enjoy the talk. Thank you, everyone. Thank you, Roger. Pleasure to be here.