Compass Therapeutics, Inc. (CMPX) Earnings Call Transcript & Summary

Good afternoon, everyone. This is Kelly McCarthy from the Morgan Stanley Healthcare team, and it's great to be here with the management team of Compass Therapeutics. I'm joined by CEO, Tim Schuetz (sic) [ Tom Schuetz ]; and CFO, Barry Shin. So thank you both for coming. Hopefully, it's been a productive conference for you so far. Before I get into the Compass Therapeutics' story, I'm just going 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.

So Tom, let's start with you. For those who may be less familiar with the Compass Therapeutics' story, can you give a little bit of company history and just how your pipeline kind of came together?

Sure. And thanks, Kelly, and thanks to Morgan Stanley for inviting us to the conference. And I would also echo the disclosure, frankly, I will be making forward-looking statements today, and I would refer anyone to our regulatory disclosures for risks associated with those. So again, thank you. So Compass, we're located in Boston, Mass. We're a monoclonal antibody development and discovery company in oncology. So we have 3 drugs in the clinic, 2 bispecific antibodies and 1 monoclonal antibody, and we have a fourth drug that we'll be filing an IND for in the next quarter. So our lead program is a DLL4 VEGF-A bispecific antibody called tovecimig. We announced earlier this year that, that drug hit the primary endpoint of overall response rate in a randomized trial. And we also announced last month that we are seeing fewer deaths in that study than we had originally projected, again, suggesting that we might be seeing a drug effect. Our next most advanced asset is a monoclonal antibody that we call 471. 471 is a next-generation CD137 agonist antibody. We presented data for that drug twice at scientific meetings last year, clinical data at ASCO and scientific data at SITC. At ASCO, we presented response data, and we had 5 responses in the post-checkpoint inhibitor patient population. Obviously, an incredibly difficult patient population to treat. We saw responses in patients with melanoma, small cell lung cancer and mesothelioma. So we can talk more about that in a minute. We also have interestingly a PD-1, PD-L1 bispecific antibody. And we set out to develop next-generation checkpoint inhibitors, and we did a screen, and that drug came out of a screen for synergy with PD-1 blockade. It could be a very important scientific discovery. But last month, we announced some preliminary data from a Phase I study, and we had very deep and very important responses in 2 patients, one with non-small cell lung cancer and one with triple-negative breast cancer. So very, very interesting early Phase I dose escalation, and we already have signals of efficacy in that study. So very, very important. And then finally, we have a PD-1 VEGF-A bispecific antibody, obviously, a well-known new class of drugs. The IND for that drug that we call 726 will go into the clinic -- will be filed next quarter.

Terrific. Well, congrats on all the progress in the pipeline. You have a deep pipeline in oncology. Maybe we can start with tovecimig and you're progressing that program initially in BTC, in biliary tract cancer. Maybe you can talk a little bit about the current treatment paradigm and the limitations there today.

Sure. So yes, we've finished a Phase I program for tovecimig. Again, DLL4 is delta-like ligand 4. That's the cell surface ligand for Notch-1. VEGF-A, of course, is the well-known soluble ligand for the VEGF family of receptors. It's the target of many important drugs, including, of course, Avastin. So in our Phase I program, we saw responses in patients with colorectal cancer and gastric cancer. And then we did a Phase Ib study where we combined tovecimig with chemotherapy. And in that study, we had a very important signal of efficacy in patients with biliary tract cancer. We had 2 patients in that study with incredibly deep and durable responses who had biliary tract cancer. We had 2 patients with 40% to 60% tumor declines, and those patients were on drug for more than a year in a setting where the median overall survival of that patient population is literally measured in single-digit weeks, believe it or not. So we confirm that in a Phase II study, shared those data with FDA and in discussions with FDA. We designed a randomized trial, which we're running now. So we're running a randomized trial of tovecimig plus paclitaxel versus paclitaxel alone in patients with advanced biliary tract cancer who have received 1 prior line of therapy. So that's a pure second-line study. And you asked a question about the current treatment paradigm. I think that's a very important question. So biliary tract cancer turns out to be an incredibly common disease. At least 25,000 patients annually in the United States are diagnosed with biliary tract cancer. There is a frontline regimen with biliary tract cancer, which is chemotherapy plus a checkpoint inhibitor, gemcitabine and cisplatin with either pembrolizumab or durvolumab. So there is a frontline regimen, but there's no labeled second-line standard of care in biliary tract cancer. So it is a massive unmet medical need with no labeled therapy. The only therapies for patients after the failure of frontline therapies are for the small fraction of patients who have an actionable mutation, which is only, unfortunately, about 15% to 20% of the patient population, either have an IDH1 mutation, an FGFR2 mutation or have a HER2 amplification. But other than that, the other 80% to 85% of patients literally have no available second-line therapy. So we're positioning the combination of tovecimig and paclitaxel to become second-line standard of care for patients with this disease.

You mentioned it at the outset, but I don't want to gloss over it. You just put out kind of updated guidance around the data from this study coming out in Q1 2026, which is an update to the timing. What does that imply? Is that encouraging? What do we...

Sure. So earlier this year, in April, we announced the results of the analysis of the primary endpoint of the study. So we had -- the primary endpoint of the study was overall response rate. We had a 17.1% response rate in the combination arm compared to 5.3% in the control arm. So standard therapies for these patients today only have about a 5% response rate. So we more than tripled that response rate, and that was statistically significant with a p-value of 0.031. At that time, we provided some guidance on when we might do the analysis of the secondary endpoints. And the secondary endpoints in the study are progression-free survival and overall survival. And those -- the analysis of those endpoints is triggered by overall survival events. So in April, we announced that we provided guidance that we might hit the event threshold by the end of Q3 and then complete the analysis and report the analysis in Q4. So we based that guidance on the number of deaths we were seeing in the study over the period of December, January, February and March. We used that to provide the guidance. Then what we saw in April, May, June, July and August was a very substantial decrease in the rate of deaths in the study. So much so that we felt like we had to update the guidance. So simply said, there are more patients alive and those patients are living longer. So just one caveat, it is pooled mortality data, so from both arms of the study, of course. But I think we know from our response rate data that the patients in the paclitaxel arm have not derived great benefit from paclitaxel. So we believe that the fact that there are more patients alive and those patients are living longer suggests that we could be seeing a treatment effect as measured by overall survival. And if that's true, then we have the potential to have a very significant drug here. And those analyses now are projected to occur in Q1 of 2026. Today, we have more than 18 months of median follow-up in the study, and we have well over 20% survival in the study. So one chemotherapy regimen that is used in this patient population, which is FOLFOX, which are 3 chemotherapy drugs, 5-FU leucovorin and oxaliplatin at 18 months in a randomized study, the overall survival with FOLFOX in this disease setting is less than 10%. So with 18 months of median follow-up, we are well over 20% and months away from reporting that analysis.

So potential for a big impact in the second line.

That's right.

That's clear. I know you have an investigator-led study out of MD Anderson that actually is looking at adding tovecimig to the first-line therapies. What do you expect to learn from this study? Do you think that there is potential as a first-line drug here?

Sure, absolutely. So that was a very interesting development, I think. So the second-line study that we're currently running, we're doing our best to stay blinded to the results from that study. We, of course, were totally blinded to the central review of overall response. So -- but the study is open label at the sites. And after MD Anderson, which was the -- MD Anderson was the second largest enroller in the study. After they had enrolled 7 or 8 patients, they sent us a proposal to add tovecimig to the frontline regimen of gemcitabine, cisplatin and durvalumab. So a really nice development. That study is underway at MD Anderson now. It's enrolling patients. We have a meeting in a couple of months with the investigators to get an update on the data from that study. And I think it's very important to think about how we might eventually move tovecimig to the frontline setting because in the second-line setting, we believe it's a very significant commercial opportunity because at least 15,000 patients annually in the United States should be eligible for this therapy, but the frontline population is about 50% more, so about 25,000 patients or so. But more importantly, patients treated in the frontline setting would be on therapy longer. So having this approved in the frontline setting would be a very significant increase in the commercial opportunity for this drug. So watching that study closely. And hopefully, we'll get some data from that study in 2026.

Okay. And then just thinking about the tovecimig opportunity beyond BTC, you've also shared you're planning a Phase II basket study in a broader set of DLL4-positive cancers. Can you talk a little bit about what tumor types might be most relevant for your exploration?

Sure. So in the first half of next year, we're going to -- we're planning to begin a Phase II study of tovecimig. And that's a basket study because we're going to enroll patients with a whole set of different tumor types. So it turns out there are many DLL4-positive malignancies. So an example, some examples, colorectal cancer, hepatocellular cancer, gastric cancer, glioblastoma, of course, a massive unmet need, ovarian cancer, renal cell cancer. So there's a whole set of indications that are DLL4 positive. Now we're developing a diagnostic assay for DLL4, and we may or may not use that for this study, depending on where we are with that. But the purpose of this study would be to do a classic Phase II signal finding study where we look for signals in different indications to help guide us in planning for post-approval studies. So if everything goes well with our progression-free survival and overall survival analyses for the biliary tract cancer study, we would follow that with an important interaction with the FDA. And of course, depending on the data and FDA's perception of the data, we could potentially file a U.S. license application next year. And then we have Fast Track status. So if we get a priority review, we could potentially get this drug approved in early 2027. And then we would really need to focus on planning for label expansion studies. And we'd like this study to guide us by providing a signal-finding opportunity for us in multiple other malignancies.

Terrific. I want to capture some of the rest of your pipeline. So I'm going to turn us to the PD-1, PD-L1 bispecific now that's 8371. And you sort of framed this program as the next generation of checkpoint inhibition and others haven't really not gone into this particular bispecific target space. So can you discuss kind of the mechanistic rationale for combining these 2 targets?

Sure. So when we first started thinking about the concept of next-generation checkpoint inhibitors, we really set out to try to make an important discovery in an area where patients really have no available therapies. So once a checkpoint inhibitor fails a patient, the post-checkpoint inhibitor patient population, there's little, if anything, available for those patients. So like many important problems, in order to address that problem, we first had to develop a novel technology. So we developed a screening tool that we call StitchMabs. StitchMabs is a tool that allows us to combine monoclonal antibody fragments together, and it allows us to do screens for synergy in bispecific antibodies. So when we set out down this pathway, we did not allow ourselves to be biased by any conceptual combination. So we did not set out to develop a PD-1, PD-L1 bispecific antibody. We did a screen for synergy with PD-1 blockade. And I think we made potentially a phenomenally important discovery, which is dual ligand and receptor blockade is between 100 and 1,000-fold more potent than either alone. It was a completely unexpected discovery. Because of that, we spent probably a year investigating the mechanism of action, and we discovered that this bispecific actually is a cell engager via PD-1 and PD-L1. It's unequivocally a cell engager. It also does something really incredible, which is it puts some torsion on PD-1, and it leads to the cleavage of PD-1 off of effector T cells. So PD-1 is the main way that T cells are suppressed. So we convert PD-1 positive T cells into functional PD-1 negative T cells. It's an incredibly novel mechanism of action. So I believe that we are on the cutting edge of defining the entire concept of next-generation checkpoint inhibition. It is superior in preclinical studies, and we took it into a Phase I study, and we're now in the middle of dose escalation.

So what have you seen from the dose escalation? You said 2 patients with deep partial responses. Can you talk a little bit more about dose escalation?

Sure. So last month, we announced an update from this study. So we're just beginning the dose escalation in this study. We're doing a typical so-called 3+3 design, where we enroll 3 patients. If there are no dose-limiting toxicities, we move to the next dosing level. So we've done 4 dose levels so far. We've seen no dose-limiting toxicities. So 4 dose levels times 3 patients, it's only 12 patients so far. We've seen no dose-limiting toxicities, and we have 2 incredible responses. CT scan images of these 2 patients are in our corporate deck on our website, and I would really encourage folks to take a look at those CT scans because they're very dramatic. We have a complete resolution of target tumors in a patient with non-small cell lung cancer who had previously seen a PD-1 blocker. So this patient also has squamous histology, which is less likely to respond to checkpoint inhibitors. This patient treated in the fourth-line setting with 6 centimeters of tumor burden had complete disappearance of his tumors. Maybe even more importantly, we have a patient with triple-negative breast cancer, 1 out of only 3 patients with triple-negative breast cancer enrolled in the study who actually relapsed while receiving pembrolizumab. So this patient is clearly refractory to checkpoint inhibitors, then got Trodelvy, then got 2 additional chemotherapy regimens. This patient was treated in the fourth-line setting with almost 9 centimeters of tumor, including a large metastatic tumor on her heart. This patient's tumors have also completely disappeared. She's gone from 87 millimeters of tumor to 7 millimeters. So a near complete response. So this patient could become a complete response in the future, obviously continues to receive therapy. So a really incredible antitumor response in these 2 patients that I think we saw with our preclinical discovery of this drug. So we announced last month that we're going to do expansions of the Phase I study in patients with non-small cell lung cancer and triple-negative breast cancer. We'll take 2 doses forward. By the end of this month, we'll have all the safety data from the fifth and final dosing cohort, and we'll make a decision about doses later this month, and we'll begin the cohort expansions in the fourth quarter. And we also hope to present the full dose escalation data at a medical meeting later this year. That abstract submission is in process. So stay tuned for that.

Okay. We will stay tuned. incredible update. We're looking forward to hearing more from that program. And then maybe we can talk a little bit about the PD-1 VEGF space and your program 726.

Sure. So scientifically, what we work on at Compass is angiogenesis inhibition, and the example of that, of course, is tovecimig. DLL4 and VEGF-A are both inhibitors of angiogenesis. And we also work on immune checkpoints. The example of that, of course, is 8371, our PD-1, PD-L1 bispecific antibody. So the overlap of that led us to explore the concept of a PD-1 VEGF-A bispecific antibody. And we worked on this concept for about a year or so, and we explored many different permutations of this. As you alluded to, I think it's now, in my opinion, anyways, I think it's established that this is a new class of drug. Ivonescimab has, I think, clearly established this as a new class of drugs. And I think even before ivonescimab, we saw the study that Roche did in hepatocellular cancer where they combined bevacizumab with atezolizumab, VEGF-A and PD-L1, and that regimen is now frontline standard of care in hepatocellular cancer. So the combination of angiogenesis inhibition and checkpoint inhibition, I think, is now a well-validated therapeutic approach. So we also explored this. We ended up with a 2x2 valency bispecific targeting VEGF-A and PD-1. In the test tube, our drug that we call 10726 is better, I put that in air quotes, it's better at blocking PD-1 signaling than ivonescimab, okay? It binds PD-1 better, and it has more potent PD-1 blockade. Because we made that observation, we decided to do head-to-head studies in mice. This was a little risky, of course. You do a head-to-head study, might be worse, but it wasn't. It's better. So in head-to-head preclinical studies in mice, 10726 is superior at controlling tumors than ivonescimab. So take that for what it is. Ivonescimab is well established. It has clinical data in multiple indications but we have some evidence that not only might we have a better drug, we clearly have a well-differentiated drug. And I think you're correct. Obviously, you're correct that this is a very crowded space now. But I always believe no matter how many drugs are approved, there's always room for better drugs. So we're wrapping up all of our IND preparations for this drug, and we'll submit the IND for 10726 in the fourth quarter of this year, which would put us in the clinic in 2026, and we should be able to deliver some preliminary clinical data next year.

Terrific. I want to spend just a minute on manufacturing because I think it sometimes does get overlooked. What steps have you taken to ensure readiness on that front?

Sure. This -- I think I agree with you fundamentally that manufacturing is commonly overlooked in these development programs. And I think we have spent a lot of time over the past 4 or 5 years, and we've developed some important bispecific manufacturing expertise at Compass. So in the early days of bispecific antibodies, many of these constructs were somewhat limited, frankly, by manufacturing. So when you think about that, you have a novel drug, but you can't make it. So beginning with 8371, our PD-1, PD-L1 bispecific and then also with tovecimig, our DLL4 VEGF-A bispecific antibody, we've focused on developing bispecific manufacturing expertise at the company. And for both of those drugs, we already have for both of those drugs, commercial-ready yields. So we have yields that are in the 5-gram per liter range, which that's just a number, might not mean much to many people, but that is better than you see with many monoclonal antibodies. Humira, for example, I believe, is in the 1 gram per liter range. So we have really incredible yields. We've applied all that technology to 10726 as well. 10726 is in the same ballpark. So all 3 of our bispecific antibodies manufacture incredibly well. And for tovecimig, we've now initiated the BLA-ready manufacturing runs that we'll need to file a license application if we get there.

Do you view the manufacturing piece as part of your differentiation?

I do. I do. And I think it is an incredibly important know-how that we have developed at the company that we're able to manufacture these bispecifics with these kinds of yields.

Okay. Barry, I want to ask you one. You just completed a very successful financing. So congratulations on the upsized deal. Can you talk about what that recent raise did for your cash runway and your ability to continue to invest across the pipeline?

Yes, absolutely. And as a bit of context, in early August, Tom provided what we thought was a pretty incredible update on the company's progress. That update included the revised time line for tovecimig, where patients were living longer than we had expected. It included the, quite frankly, astounding and unexpected data for 8371, the PD-1, PD-L1 bispecific, where 2 patients showed 100%, 90% target tumor reductions. And then also, we updated on the progress of 10726, the VEGF PD-1 bispecific that showed superiority to ivonescimab. So with that, we thought it was a very exciting update and investors agreed. With that investor support, we catalyzed a $138 million upsized financing that we close in mid-August. That extends our runway into 2028 and importantly, extends the runway through a whole slew of potentially transformational milestones. For tovecimig, we expect the PFS and OS data in Q1 of 2026. And again, that is potentially transformational for the company. And then BLA filing in mid-2026 and potential approval shortly after. For 471, it funds the DLL4 -- sorry, the basket study, NCAM positive basket study. For 8371 as well, it funds the cohort expansion study that we're looking forward to. And then for 726, it takes the asset into the clinic and ideally proof-of-concept data in 2026. So a whole lot of milestones coming up, and we're on solid financial position to execute on all of them.

You guys have a very busy catalyst calendar coming up. I guess what do you think is resonating the most with investors about the story? And what are you 2 personally most excited about? If we were to be talking again on this stage in 12 months, what do you want to look back and say, we nailed that?

Yes. I'll start with what's resonating with investors. Clearly, tovecimig, the progress with tovecimig, I think the financing both validates the existing foundation of clinical data that we have and then also the prospects for the upcoming data readouts. So investors are clearly excited on that -- about that and focused on that readout. Now I do think that we are getting significantly more interest on 726. And then even today on 8371, again, this was an unexpected outcome and is just now getting on to investors' radars. So investors focused on tovecimig. In my personal view, I think we're being valued and possibly undervalued on tovecimig alone. But I do hope that once we do show the, knock on wood, positive PFS OS data that we start to get additional value built into the company for our exciting pipeline. Tom, I'm sure you have other thoughts as well.

Yes. I think we have an opportunity here to develop drugs for patients that have no therapies, right? That is an unbelievable and incredibly important effort. And I always end up -- you said what's personally important to me, right? I appreciate that question. And I always think about individual people, right? And think about this patient with triple-negative breast cancer, okay? She feels a lump in her breast. She goes to a doctor, has a biopsy. Doctor says, "ma'am, this is a bad disease. We have to go all in on chemotherapy and KEYTRUDA." She gets that in the neoadjuvant setting before surgery, then she has surgery, a lumpectomy. Then she has more KEYTRUDA and she develops metastatic disease while receiving KEYTRUDA, one of the best drugs ever developed. And then her doctor says, "okay, now we have a real problem." Then she gets Trodelvy and doesn't respond to Trodelvy. Now the doctor says, "Okay things are not looking good here." She gets chemotherapy, doesn't work. She gets another chemotherapy, doesn't work. Now she's in the doctor's office and the doctor is looking at her and saying, "we've got a real problem here. I'm sorry, but we have 2 choices. You can either go into hospice. You probably have 3 to 5 months to live or you can go into a Phase I trial and receive a drug that nobody has ever received." And she says, "all right, I'll do that." And now her tumors are gone. This is somebody who was making a decision about going into hospice for God's sake. So think about what that means for that patient. And that's what this means to me personally.

I could not have asked you to end on a better note. Unbelievable updates across your whole pipeline. And I think the anecdotes really bring to life what you're doing for patients. So we are going to stay tuned, and we're looking forward to your upcoming updates. And thank you for joining us at the Morgan Stanley Healthcare Conference. Great having you here.

Great. Thank you so much.

Great. Thanks, Kelly.

Thank you, Tom. Thank you, Barry.