CytomX Therapeutics, Inc. (CTMX) Earnings Call Transcript & Summary

Good morning, and good afternoon, everybody. Thank you all for joining. My name is Michael Klem. I'm a member of JPMorgan's health care investment banking team here. It's my pleasure to introduce Sean McCarthy, the CEO of CytomX Therapeutics. [Operator Instructions] We also have Amy and Carlos here for Q&A after. But without further ado, I'll turn it over to you, Sean. Thanks.

Great. Thank you very much, Michael, and it's a pleasure to be here today. Thanks to the organizers for the opportunity to present an update on our company. It's a real pleasure to provide an overview of our recent progress on our conditionally activated antibody therapeutic programs for the treatment of cancer and our plans for 2021 and beyond. If we could advance -- if I could draw your attention to Slide 3. Before I begin, let me remind you that I will be making certain forward-looking statements during this presentation. Please refer to our regulatory filings with the SEC that outline key risks and uncertainties for our business. Moving to Slide 4. CytomX is an innovative leader in the field of conditionally active therapeutic antibodies that leverage our deep understanding of tumor biology. Our goal is to create antibodies that are selectively activated in the tumor microenvironment and less active in healthy tissues, thereby addressing targets previously considered undruggable because of their presence on normal tissues. Our 2 lead product candidates, CX-2009 and CX-2029, are conditional antibody drug conjugate is currently in Phase II in areas of significant unmet medical need. We also were working towards additional IND filings for new product candidates this year. Our industry-leading technology has been successfully partnered with several multinational pharma companies including Bristol-Myers Squibb, AbbVie, Amgen and Astellas. These partnerships are allowing us to broaden the application of our technology under an important source of non-diluted financing, resulting in a maintained strong balance sheet with $321 million of cash as of the end of the third quarter of 2020. Moving to Slide 5. I am delighted to work with a terrific team of experienced executives with wide and varied backgrounds in our industry. Together, this team brings decades of experience across research, drug discovery, drug development, financing and business development, and together, we're really excited about the year ahead. Moving to Slide 6. At CytomX, we're advancing a broad and deep clinical pipeline, a highly innovative pipeline that includes 3 conditional antibody drug conjugates, 2 conditional or checkpoint inhibitors and our first conditional bispecifics T-cell engager. CX-2009 is a wholly-owned conditional ADC targeting the novel tumor antigen CD166. We're currently evaluating CX-2009 in a 3-arm Phase II study focused on breast cancer. CX-2029, our anti-CD71 conditional ADC, partnered with AbbVie and a global co-development alliance, is currently being evaluated in Phase II expansion cohorts in 4 different tumor types. Based on our current enrollment projections, we anticipate initial data from our CX-2009 and CX-2029 Phase II studies to be available in the fourth quarter of this year. Our third conditional ADC is CX-2043, targeting EpCAM or TROP-1, which we plan to bring to an IND filing later this year. In addition to our conditional ADCs, we also have 2 conditional checkpoint inhibitors targeting CTLA-4 in clinical studies in partnership with Bristol-Myers Squibb. One of these is in a randomized Phase II study in melanoma, and the second is in Phase I. The leading edge of our collaboration with Amgen is CX-904, our first conditional T-cell engaging bispecific antibody targeting CD3 on T-cells and EGFR on tumor cells, and we're progressing CX-904 towards potential IND filing this year as well. Moving to Slide 7. Our unique approach to conditional activation of therapeutic antibodies is called the Probody platform. Our technology has been under development for more than a decade and is applicable to multiple antibody formats, including ADCs, checkpoint inhibitors and bispecifics. Probodies are designed to be inactive before they reach tumor tissue. Each antibody is genetically engineered into a Probody with a unique masking peptide that limits its ability to bind to target. Enzymes called proteases in the tumor microenvironment unmask the Probody, allowing it to bind to target on tumor cells. Importantly, these proteases are not widely active in healthy tissues, making it possible for our antibodies to selectively target tumors while sparing collateral damage to normal tissues and normal cells. Fundamentally, our platform is designed to create a therapeutic window for previously undruggable targets or to improve therapeutic window where there is room for improvement, for example, with checkpoint inhibitors. Moving to Slide 8. I think we all agree that after many years of effort, the broader field of ADCs has clearly experienced a renaissance at the moment with several exciting new drug approvals and major transactions occurring in 2020. We've been active in the ADC field for some time and conditional ADCs that we also call Probody drug conjugates, or PDCs, are at the forefront of our clinical pipeline. Moving to Slide 9. Although we're seeing great progress in the ADC build, a significant challenge is posed by a dearth of new targets. This is because the ideal target for a conventional ADC should be high in tumor, but low in normal tissues so as to avoid unexpected target-mediated toxicities. Probody drug conjugates expand the ADC target landscape by enabling us to address highly expressed tumor antigens regardless of their presence on normal tissues by employing our masking technology. In this way, we can convert the undruggable to the druggable, and by definition, create new and potentially first-in-class therapeutics. I'd now like to move to a discussion of CX-2009 genetic name praluzatamab ravtansine, our conditional anti CD166 ADC, which is in Phase II for HER2 non-amplified breast cancer. So let's now move to Slide 11. Despite recent progress in new breast cancer drug approvals, a substantial unmet need remains, especially in the metastatic setting. And we need more options with single-agent activity in late lines of treatment. CD166 is broadly and highly expressed in breast cancer, including in HER2 non-amplified tumors, and our development program is currently focused on this category. Moving to Slide 12. High CD166 expression in breast cancer, and indeed, in many other tumor types, would make this abundant cell surface antigen an attractive target for a conventional ADC, if it was not for the fact that it's also highly expressed in normal tissues, posing a significant risk to clinical development. We reasoned, however, that a conditional ADC to CD166 could unlock potential in this target by targeting the ADC activity towards CD166 on tumor cells and away from healthy tissues. Accordingly, we designed CX-2009 shown on this slide, a Probody drug conjugate utilizing the maytansinoid payload DM4, a well-characterized microtubule vector. Our therapeutic strategy in our Phase I work -- our Phase I clinical work was to evaluate the tolerability of CX-2009, the doses in the anticipated therapeutic range of the payload, and of course, to look for initial signs of clinical activity. We use DM4 since its toxicities are well-described, particularly reversible ocular events, providing us with a precedented background on which to evaluate any on-target toxicities that we may have seen from CD166. Slide 13 shows the patient demographics for this study. We have previously reported our Phase I dose escalation study in solid tumors and the manuscript is in preparation. I would like to focus here on our experience in breast cancer, one of several indications in which we saw promising signs of clinical activity in Phase I. 39 patients with HER2 non-amplified breast cancer were enrolled into the Phase I study. 28 of these patients have hormone receptor-positive breast cancer and 11 patients had triple-negative breast cancer. These patients are heavily pretreated with 8 and 7 median prior regiments, respectively. The majority of hormone receptor positive patients had tumors with high CD166 expression, whereas about half of the patients with triple-negative showed high CD166. And that's shown on the right-hand side of this slide. Moving to Slide 14. Data are presented here from an August 2020 data cutoff, and these data were presented in full at the 2020 San Antonio Breast Cancer Conference. Among 32 evaluable patients treated with 4 mg per kg or more of CX-2009, 3 unconfirmed partial responses were observed in evaluable patients with TNBC, and 2 confirmed partial responses were observed with patients with hormone receptor positive HER2 non-amplified breast cancer. Most responses occurred early, as seen in the slide -- on the right of this slide, although one patient reached a partial response after 40 weeks of treatment. Another patient treated initially with 6 mg per kg every 2 weeks demonstrated a deep response that was maintained even after dose reduction at week 8 to 4 mg per kg every 2 weeks. This response ultimately converted to a complete response in target lesions at 44 weeks, and this patient was still on treatment as of the data cutoff. The overall clinical benefit rate at 24 weeks of these 32 heavily pretreated patients was 28%, with several patients achieving disease control beyond 6 months. Shown on the next slide, on Slide 15, is a case study of a 41-year old patient with TNBC, who progressed on prior pembro-paclitaxel, and then actually also progressed on sacituzumab govitecan. After receiving CX-2009 at 8 mg per kg, the patient responded with a 48% reduction in target lesions, indicating a lack of cross resistance to these prior therapies. This response was unfortunately not confirmed due to dose interruption or an ocular adverse event, which is an expected toxicity from the DM4 payload, for which prevention measures were not implemented in Phase I. Nonetheless, this data shows that CX-2009 can demonstrate clinical activity in heavily pretreated patients with TNBC. Slide 16 shows the tolerability of CX-2009. Adverse events from the Phase I study are summarized here. Toxicities appear to be consistent with the DM4 payload and were ocular neuropathic and hepatic in nature. No dose-limiting toxicities were observed up to a dose of 7 mg per kg administered every 3 weeks, and this was selected as the recommended Phase II dose for further study. The design of our Phase II study is outlined on Slide 17. We have now launched a 3-arm Phase II of CX-2009 in breast cancer. Arm A is enrolling patients with hormone receptor positive HER2 non-amplified breast cancer, and Arm B is enrolling patients with TNBC. Arms A and Arms B are evaluating monotherapy. Arm C is evaluating a combination of CX-2009 with CX-072, our proprietary anti-PD-L1 Probody, also now known as pacmilimab. And this is in patients that are positive for both PD-L1 and CD166. We've previously reported single-agent activity for CX-072 pacmilimab in triple-negative breast cancer and other tumor types, together with a favorable tolerability profile for this novel checkpoint inhibitor that renders it an attractive combination partner to CX-2009 and potentially other agents in our pipeline as we move forward. Importantly, prophylactic treatment for ocular toxicities is mandated in this Phase II study. And based on our work and that of others, including ImmunoGen, which is the originator of DM4, we are optimistic that this toxicity can be effectively managed in the Phase II setting. We expect initial results from Arms A and B to be available in the fourth quarter of 2021, and of course, subject to any changes from unforeseen impact of the pandemic. Arm C, due to the requirement for selection of patients with both PD-L1 and CD166, is expected to enroll a little bit more slowly. So we're probably looking at data emerging from this arm in the first half of 2022. I'd like to move now to a discussion of our second lead product candidate, CX-2029, which is another conditionally active ADC targeting, in this case, CD71. So moving to Slide 19, some background on CD71. CD71 is a transmembrane glycoprotein that efficiently internalizes iron-bound transferrin. It internalizes into dividing cells, and it's one of the most efficient [indiscernible] transport systems that we know of. CD71 has been known for decades to be highly expressed on malignant cells in many different tumor types due to their active metabolism and their increased requirements for iron. CD71 is also known to quickly and efficiently internalize antibody drug conjugates, as shown here in the upper right panel of this slide, and this makes it an ideal target for targeting cytotoxic payloads into cancer cells. However, CD71, as you might imagine, given its role in iron metabolism, is present on all healthy dividing cells, particularly those with a high iron requirement such as the bone. In this regard, CD71 can be considered perhaps the quintessential undruggable target in the ADC space. So the question we've asked ourselves is, whether we can unlock the potential of CD71 with our Probody technology. So moving to Slide 20. To answer this question, we designed CX-2029 with our Probody platform, and this is a conditional anti-CD71 ADC conjugated with the auristatin payload MMAE. CX-2029 shows potent anticancer effects in animal models, as shown here on the right of this slide in a head and neck model, for example. We've actually seen that activity in dozens of tumor types in preclinical studies. So this is a broadly active molecule. Now we know from our preclinical work that an unmasked conventional antibody drug conjugate to CD71 is lethally toxic in animals at subtherapeutic doses. So an unmasked ADC has 0 therapeutic window. The toxicities observed are principally hematologic in nature, arising from the payload, and also likely the role that the target plays in the bone marrow -- and in bone marrow biology. So our strategy here with the masked ADC was to evaluate in the clinic, whether via masking the antibody, we could achieve therapeutic levels of a conditional ADC to CD71 in cancer patients. Now we know from our work and that of others that to achieve therapeutic effects in cancer patients with CX-2029 at the MMAE payload, we would likely need to get to doses in the 2 to 4 mg per kg range. Moving to Slide 21. Our Phase I dose escalation enrolled 45 patients over 8 dose levels. Patients were not selected based on CD71 expression, and the immediate number of prior treatments was 3. Slide 22 shows encouraging evidence of anticancer activity that we observed at 2 mg per kg and above. So we were able to get fairly quickly in the dose escalation to therapeutic levels of this novel direct candidate in patients, which is very important. The data presented here are from an August 2020 data cutoff, and interestingly, the activity that was observed principally in tumors with squamous histology, with confirmed partial responses being seen in 2 patients with squamous non-small cell lung cancer and in one patient with head and neck cancer. Moving to Slide 23. Let's take a closer look at the activity in squamous tumors. 4 squamous non-small cell lung cancer patients were treated in this study. Of the 3 squamous lung patients completed at 2 mg per kg or above, 2 experienced partial responses that were confirmed, and the third patient had stable disease. The fourth lung patient, the fourth squamous lung patient in the study, was treated at 1 mg per kg, and that's a dose that thought would be predicted to be subtherapeutic. And indeed, this patient did not, unfortunately, respond. There were 8 head and neck squamous cell carcinoma patients enrolled and treated with doses of 2 or 3 mg per kg in the study. One of these patients experienced a confirmed PR, and 6 of them had stable disease. The 8 patients with head and neck cancer did not have a post baseline scan. So this is very encouraging activity. We see this as a clear signal of activity for CX-2029 in squamous tumors. And the spider plot on the right of this slide shows early but promising evidence as to the tolerability of CX-2029 and the durability of response. Slide 24 shows a case study of a 66-year-old individual with metastatic head and neck cancer. And specifically, this is a nasopharyngeal carcinoma. This patient was treated with 3 mg per kg of CX-2029. This patient had received multiple treatments prior to study entry, including chemo and radiation, and upon progression, had also received a pembro-based regimen. The bottom panel shows the decrease in one of this patient's target lesions in the liver, a fairly large lesion. And at week 8, the sum of all target regions in this patient have decreased by 43%. A PR that was confirmed at the subsequent scan was sustained through the last scan, preceding the data cutoff date several months later, demonstrating clearly that CX-2029 can elicit antitumor effects. Turning now on Slide 25 to safety. CX-2029 was generally well tolerated. The most frequently observed adverse events were hematologic in nature. The most commonly occurring grade 3 or higher adverse event was anemia, recurring in 49% of the 45 patients treated across all dose levels. We continue to investigate the etiology of the anemia, which may arise from payload effects on the bone marrow, and may also reflect the important role that CD71 plays in novel red blood cell development, and we're investigating this in more detail. Patients with hematologic side effects were managed typically with red blood cell transfusions or with supportive care and/or supportive care. So taken together, these first clinical results for CX-2029 show us that we have succeeded in opening a therapeutic window for CD71 for the first time, a previously undruggable target with enormous potential across multiple tumor types. And we're now really excited to be moving 2029 into the Phase II setting. So encouraged by these findings. We're pushing ahead with Phase II expansions, evaluating the dose of 3 mg per kg in patients with non-small cell lung, squamous non-small cell and squamous head and neck because, as you might imagine, based on the Phase I data. We're also enrolling patients with esophageal and gastroesophageal junction cancers and also DLBCL. And these are indications in which we've seen, among others, promising preclinical activity. This is on Slide 26, by the way, the study design for the Phase II expansions. As with CX-2009, we are working towards initial results in this expansion phase to be available in the fourth quarter of 2021, again, subject to any limitations on enrollment due to the pandemic. I'd like to stay with the theme of drug conjugates, but now switch to our third conditional ADC, CX-2043, which is directed against another really high potential, but again, previously undruggable target EpCAM, also known as TROP-1. So moving to Slide 28. EpCAM is an epithelial cell marker that has been known for a long time to be expressed on many solid tumors. EpCAM has the molecular properties to internalize ADCs and to deliver cytotoxic payloads into cells. But its wide expression on normal epithelial tissues has precluded the development of systemic anti-EpCAM therapies to date. EpCAM has been -- has in fact, been successfully targeted, though, with locally delivered agents. For example, agents that are instilled into the bladder, and this provides a level of validation for the target. So this is really a very interesting target if we can find a way to actually drug it. So CX-2043 is our conditional ADC, again a Probody drug conjugate designed for systemic administration. CX-2043 was designed and constructed in collaboration with ImmunoGen, using a novel payload, DM21, which offers increased potency, enhanced stability, and also improved bystander effects. We recently presented comprehensive preclinical data for CX-2043 at the AACR-EORTC triple meeting in Q4 last year, and we're targeting IND filing by the end of 2021. We acquired full commercial rights to CX-2043 from Immunogen in 2019. Now I'd like to move to our alliances on Slide 30. As we build our company and our pipeline, we continue to work closely with our major partners, BMS, AbbVie, Amgen and Astellas, with whom we're advancing multiple Probody programs. In collaboration with BMS, we continue to work to develop potentially safer and more effective versions of ipilimumab, the anti-CTLA-4 checkpoint inhibitor. BMS-986249 is a Probody version of ipilimumab that has shown an attractive tolerability profile as monotherapy and in combination with nivolumab in Phase I studies in patients with solid tumors. These data were presented by BMS at ASCO in 2020. BMS is now conducting a randomized study, evaluating the tolerability and activity of BMS-986249 in patients with metastatic melanoma in combination with nivo. And importantly, this study is a control study looking at Probody ipi plus nivo compared to ipi-nivo. And again, this is in the frontline melanoma setting. So a very exciting study, an important study being executed by Bristol-Myers at the moment. Now BMS-986288, which is a Probody version of a non-fucosylated ipi, is a more potent version of ipi that we've turned into a Probody, is also being advanced by BMS in this partnership, and that program is in Phase I. This program is designed to broaden the therapeutic window of a more potent version of ipi. Phase I study enrollment is continuing by BMS, and we look forward to additional progress with BMS on both of these programs as we move through 2021 and beyond. Our collaboration with AbbVie is centered on the CX-2029 program, for which CytomX retains 35% of U.S. commercial rights. We also continued drug discovery efforts with AbbVie on earlier-stage programs that leverage our Probody platform. We also now have 2 alliances in the exciting and emerging field of conditional T-cell engaging antibodies. We call them pro TCBs. With Amgen, we continue to advance our first program, which is EGFR-CD3 towards the clinic, and in 2020, we entered into a multi-target alliance, a new alliance with Astellas, also in the bispecifics space, for which we received an upfront payment of $80 million. These alliances have, in addition to -- just moving to Slide 31, in addition to providing an increased number of programs and shots on goal for our platform, the alliances have continued to provide us with significant non-dilutive financing, which is contributing to our maintained strong balance sheet. In 2020, in fact, we generated $130 million of incoming cash flow milestones under existing deals and from new business development. End of Q3 cash in 2020 was $321 million, which funds us through 2022. And this projection does not take into account future milestones, new business development transactions, or of course, additional equity finances. Moving to Slide 32. 2020 was a year of very -- a large number of important achievements for CytomX as we extended our leadership with our Probody platform. We have defined this field. We have 5 programs in the clinic, a multi-modality platform that continues to generate novel clinical candidates against undruggable targets. And with strong partnerships, we're well positioned as we enter 2021, and we're highly focused on continued execution. Our key priorities for the year ahead are to enroll our Phase II studies for CX-2009 and CX-2029 with the goal of initial data in Q4. We're also working towards IND filings for CX-2043 and CX-904, and we look forward to additional progress in our alliances. So closing on Slide 33, I'd like to again emphasize that we continue to show leadership in the emerging field of conditional antibody activation, which offers many new opportunities for the treatment of cancer. Our innovative science and platform has allowed us to build a broad and deep clinical pipeline that continues to advance as we build towards our vision of becoming a long-term multiproduct commercial stage organization that can make a significant difference in the lives of cancer patients. So thank you very much for your attention today. I believe we'll now move to Q&A. Thanks for your time.

And for the Q&A portion, I will be joined by Amy Peterson, Chief Development Officer. Hi, Amy. And Carlos Campoy, CFO. Hi, Carlos.

Hello.

So I think we're going to moderate our Q&A here. And let me see. First question is, our presentation is focusing on the antibody, but other ADC companies highlight unique linkers or toxins or payloads. Does CytomX have the optimal [ exotoxin ], or is more needed here. Yes, that's a really good question. So let me make a couple of comments on payload selection for our first programs and where we plan to go next. Given that we're blazing a trail with new targets, CD166 and CD71, we very purposely selected through our first 2 clinical programs, payloads for which we understood the clinical profiles and the clinical toxicity. So DM4, for which the principal toxicity is indeed ocular, also neuropathic and hepatitic; and MMAE, where there's neuropathy and hematologic tox. And the reason we did that is that we really didn't know, of course, going into the clinic with these ambitious programs, what kind of on-target toxicities we may see. And we wanted to be able to tease apart payload tox from target toxicity. I should also say, of course, that we're masking the antibody, but we're not masking the payloads. So with the Probody drug conjugate, we always expect to get to payload toxicities because, of course, we want to dose to the MTD. And for these chemotherapeutic agents, you really need to do that to see anticancer activity. So the toxicities that we've seen so far are expected, anticipated, and we've been encouraged by the relative lack of on-target toxicities that we've seen with these first PDCs that we've put into patients. Now as you've seen with our EpCAM program, we are beginning now to branch out into less precedented next-generation payloads, DM21 on the EpCAM Probody drug conjugate. We also acquired from Astellas, a couple of years ago, a novel payload, a site-specific conjugation strategy and novel payload chemistry, which we're also optimizing and that we will likely be using on future ADCs as well. And then, of course, we continue to, like everybody else, look for other opportunities through various licensing and acquisition strategies. So I hope that helps answer that question. Any other questions? Okay. Next question. What is the TROP-1 market opportunity? And how are these patients currently treated? How selective is this market for cancer cells? What clinical updates should we expect in 2021. Let me talk about the TROP-1 targets. This is EpCAM. So EpCAM is also on is TROP-1. We -- I would direct you to our presentation at the triple meeting last year, where we show on our poster an expression chart of the expression of EpCAM TROP-1 on solid tumors. And the short story is it's highly expressed on most epithelial tumors. So it's really been thought of as a great cancer target for a long time, but it's not been possible to develop a systemic therapy. So all I would really say is that the market opportunity is likely very large in many different tumor types. We aim to file the IND this year. I would not expect any clinical data from the EpCAM program in 2021. Amy, you may want to comment just more broadly on availability of data across the pipeline in 2021 just to reiterate that.

Yes, of course. I'm happy to do so. So we are looking forward to incoming data from a variety of expansion cohorts and Phase II cohorts that we've initiated. And so with 2009, we're looking forward to seeing data from the first 2 cohorts that Sean walked you through, cohorts A and B, which is monotherapy in the hormone receptor positive HER2 non-amplified subset of patients, which represents more than 60% of patients, and then as well monotherapy in triple-negative breast cancer patients. For the 2029 program, we're optimistic that we should have preliminary data from the first 2 cohorts, the squamous lung and the head and neck squamous. We're keeping our fingers crossed for the third cohort, which is the esophageal GEJ cohort. We think that the diffuse large B-cell lymphoma cohort might be a little challenging to enroll due to the competition that exists in this particular indication as well as the rarity of these patients. So we won't probably have that by the end of Q4 2021.

Great. Thank you, Amy. Just one other component of that question which I didn't address, was how selective is TROP-1 for cancer cells. And this is an important point because it's actually not particularly selective for cancer cells because of its expression on normal tissues, which is where our technology comes in. This is why it's been a difficult target to drug so far because it's present on epithelial structures, as the name of the target would suggest: epithelial cell molecule. So it's our technology that we're using to open a therapeutic window. But if you look at our preclinical data, it details in some -- shows in some detail the work that we've been able to do by masking the antibody to open a therapeutic window preclinically that will take it to the clinic and explore in cancer patients. We have a third question, which is for 2029, some head and neck squamous tumors seem to respond better than others. Any thoughts? And are you selecting the patients in this trial? I'll hand that one over to Amy.

Thanks, Sean. So that is almost always the case when you are evaluating activity of your drug. And given that this is a heterogeneous -- despite its all head and neck squamous cell carcinoma, it's heterogeneous in terms of the prior number of therapies that these patients received. So the fact that of all 7 patients that we could evaluate, their response was stable disease or better, that actually is really encouraging to us as we move forward now into a larger expansion. With regard to the question on selection, I think we are always interested in ensuring that we are identifying the right patients for the right drugs. We are going to be evaluating various ways in which you can measure CD71 expression to determine whether or not that can improve selection for patients. There are a variety of other things that we could be looking at. So it's target expression. It's not lost upon us that the signal we observed in this Phase I experience was in the squamous histologies. We're looking to see whether or not there might be something there that is not only related to potential sensitivity to a CD71 targeting agent, not only sensitivity to an MMAE payload, but potentially, there might be something to do with the tumor microenvironment, the protease biology that may make them more amenable to our particular platform. So we'll be investigating all of these things through the course of the expansion cohorts and hope to find something that helps us really hone in on the right patients.

Great. Thank you, Amy. Question 4. For CX-2009, it seems there are some super responders in the current trial in breast cancer. How would you explain that? Could it be higher expression of the target? And we have about 3 minutes left, I believe. So one comment I would make, and then I'll hand over to Amy, is that as I mentioned in my comments, these patients in this study were very heavily pretreated. So it's 7 to 8 median prior lines of therapy. So the conclusions we can draw from the Phase I are relatively limited. That said, we are taking some steps on target selection in the Phase II, but Amy will talk about our philosophy there.

That's right. So the 2 things that we're doing to even further enrich for these super responders, thank you for noting that, would be that we're limiting the prior therapies that these patients can have. So for the hormone receptor positive, 0 to 2 prior cytotoxics. If you recall, that these patients had a median of 8 prior therapies in our Phase I dose escalation study. So we're really focusing now on a population that we think has a higher likelihood to respond, but as well, could represent a registrational patient population as we think about how do we move these targets forward. Same thing in triple-negative breast cancer, the median number of prior therapies in the Phase I is 7. We're limiting the prior therapies. They must have had at least one but no more than 3 prior therapies, and they could have had, but are not required to have had sacituzumab. As it regards expression of CD166 and whether or not that's going to be helpful, you saw that we're selecting for CD166 expression in the triple-negative breast cancer patients. We're not selecting in the hormone positive patients. That's because of the data that you saw on the demographic slide, where all but one patient had high expression, at least by this assay. So again, just like we're doing for 2029, we will be collecting tissue and looking at whether or not there are, again, ways to improve the selection so that we can decrease the denominator, improve the numerator and benefit patients overall.

Great. Thanks, Amy. And I believe we are just about out of time, and I don't see any additional questions. So I think we'll wrap up. Thank you all for your time today. It's been great to be able to share our story and our 2021 plans. We're excited about the year ahead. And please feel free to reach out directly with further questions offline.

Thank you.

Thank you.