ALX Oncology Holdings Inc. (ALXO) Q1 2026 Earnings Report, Transcript and Summary

Greetings, and welcome to the ALX Oncology First Quarter 2026 Financial Results Conference and webcast. [Operator Instructions] Please note, this conference is being recorded. I will now turn the conference over to your host, Jason Lettmann, CEO. Please go ahead, sir.

Thanks, everyone, and welcome to our Q1 2026 results. I really appreciate you all spending time with us this morning and are looking forward to sharing these updates with you. On Slide 2 and before we start our presentation as housekeeping, here are our forward-looking statements for your review. On Slide 3, here's the agenda and plan for today. We're going to be providing an update on our key accomplishments in the first quarter of 2026. Most notably, we are very excited to share with you the data set that was presented yesterday in Munich at ESMO Breast 2026 from the analysis of our trial evaluating evorpacept and zanidatamab, which clearly showed again that CD47 expression is a key predictive biomarker for increasing durable clinical response in heavily pretreated HER2-positive patients. This further validates the results we saw with Evo and HER2-positive gastric cancer from our ASPEN-6 trial. We are also very excited today to be joined by Dr. Sara Hurvitz from the Fred Hutch Cancer Center at the University of Washington, who is a well-recognized expert in metastatic breast cancer. She will be presenting the ESMO breast data in detail and also share her views on evorpacept's significant potential within the current treatment paradigm for HER2-positive metastatic breast cancer. Our CMO, Barb Klencke, will then recap our findings to date with Evo and HER2-positive cancers across now two different independent trials and indications, showing that patients with high CD47 expression derive the greatest benefit across all key efficacy markers, including overall response rates, duration of response and PFS versus those with low expression. These results support our targeted oncology approach with Evo, notably our ongoing ASPEN-09 breast cancer Phase II trial. Barb will then provide an update from our ongoing Phase I trial with our novel EGFR-targeted ADC, ALX2004, which also continues to track to plan, and we are enthusiastic about its potential to also change care in EGFR-expressing cancers. So on Slide 4, we continue to advance both our programs, Evo and ALX2004 and are pleased with the progress on both fronts over the course of Q1 and remain well positioned to report on significant data readouts from these programs in the coming 12 to 18 months. On the Evo front, CD47, as you know, is incredibly important in oncology and has been a very difficult target to crack, yet it plays a fundamental role in tumor evasion and resistance. We've been happy to see Evo's differentiated approach to this target further validated in the clinic over now 5 different data sets, including randomized data. Further, the new data from this week's ESMO meeting and over the last year support CD47 as a strong and predictive biomarker. These new findings further support the drug is working as designed and also enables a targeted strategy, which allows us to focus on the patients who will benefit most. On ALX2004, as you know, it is a highly differentiated EGFR-targeted ADC, which was developed in-house by ALX and has the potential to be a first-in-class EGFR-targeted ADC. EGFR has also been a difficult target for ADCs historically, and ALX2004 was designed to address this with a novel epitope and linker payload construct. We have been pleased to see the strong preclinical data to date translate to the clinic and with the continued strong progress in dose escalation, where we have been able to further increase dose. Overall, both programs remain on track and well positioned. Now on Slide 4, Q1 was a quarter of continued good execution and additional positive data as well as completion of key hires and a strong financing. As you'll hear more about shortly, this week's data validate again the power of CD47 in a late-stage patient population. Here, we observed that all patients who were confirmed HER2-positive and CD47 high achieved a response, including one complete response. We also yet again saw the durability that one hopes to see with an IO mechanism with a median duration of response of 20 months and a median PFS of 22 months, which is clearly far improved over the benchmarks. While the focus for today will be on this new data set with Evo, ALX2004 is also progressing through dose escalation very well and remains on track for initial safety readout in the second half of this year. We remain as excited here about its potential as we do with Evo. Further, the financing from February, which was anchored by very strong investors, further strengthens the balance sheet and enables us to execute through these important data milestones. And last but certainly not least, we added Jeff Knight, who is now our Chief Development and Chief Operating Officer, which really further strengthens our leadership team. Jeff, as you may know, is one of the best operators in biotech in terms of building companies at our stage through the next phase, including commercialization. On Slide 6, briefly, this just highlights the execution, clinical development progress and time lines, which remain on track for both programs. Our evorpacept Phase II breast trial continues to progress well with strong enrollment globally and increasing site enthusiasm. Overall, clinical time lines remain on track as we're expected to read out 80 patients mid next year. ALX2004, as I mentioned, is also progressing well with enrollment and dose escalation ramping and are on track there for an initial safety readout in the second half of this year. Our goal and vision for both of these programs also remains unchanged as we aim to have both programs ready for registrational studies by end of next year. On Slide 7 and my last slide, here is also a snapshot of our clinical pipeline. As you know, we're advancing 2 novel cancer treatments with key near-term catalysts. And again, ASPEN-Breast is on track. And we continue to also be encouraged by our ongoing study with our partner, Sanofi, testing Evo in multiple myeloma. Our two studies in HER2-positive cancers have now been completed as we'll focus on today and ALX2004 also remains on track. I'd like to now turn the call to Barb, who will provide a quick reminder of our MOA, which is very important as we're clearly seeing it translate to the clinic, and we'll then turn the call over to Dr. Hurvitz to review the new data. Barb?

Thank you, Jason. These next few slides describe evorpacept's mechanism of action. What's so compelling to me is that the mechanism of action is entirely consistent with what we see in the clinical trials. So let me just go through this. Cancer cells commonly upregulate CD47 expression to evade immune detection. Evorpacept blocks the CD47 signaling, but it does so without an active Fc domain. So evorpacept by itself does not engage the phagocytic activity of a macrophage. We need to combine evorpacept with an anticancer antibody. And then the combination together strongly augments the ADCP activity of the antibody. And that's likely what results in the efficacy that we see in the studies that we'll talk about today. Essentially, the antibody stimulates the eat-me signal, and the evorpacept inhibits the don't eat me signal. Therefore, the combinations of evorpacept with an antibody are selectively enhancing or augmenting macrophage phagocytosis of cells that express HER2, for example, in the case of a HER2-targeted antibody while also blocking the don't eat me signal. On the next slide, we see how this mechanism of action translates into safety. Our clinical safety database is now more than 800 evorpacept-treated patients. And it also -- the clinical safety data is also entirely consistent with this mechanism of action that we show here. Evorpacept was designed again without an active Fc domain. That's shown on the right side of the slide. So when we block the don't eat me signal on normal healthy cells, such as red blood cells, there is no macrophage activation, and we avoid on-target toxicities. This is really fundamentally different than what's been tried in the past by conventional CD47 inhibitors, which have essentially failed in the clinic. They attempted to use CD47 as a tumor-associated antigen with both an active Fc domain to activate macrophage at the same time of blocking CD47. That ended up being an indiscriminate approach because CD47 is so widely expressed on healthy cells, especially hematologic cells in particular. And that led to the potential for significant on-target toxicity. So with that description of evorpacept, its mechanism of action and its differentiation, let me now turn the conversation over to Dr. Sara Hurvitz. On Slide 12, you can see her background here. She is a steering committee member for our ASPEN-09 Phase II study of evorpacept in HER2-positive breast cancer patients. She was an investigator on the evorpacept and zanidatamab trial that we'll discuss today. And she's a Senior Vice President, a Professor of Clinical Research and the Director of the Clinical Research Division at the Fred Hutch Cancer Center at the University of Washington in Seattle, where she's also the Head of the Division of Hematology and Oncology and the Smith Family and Dow Chair of Women's Health. She's an expert in the management of breast cancer and in the development of novel targeted therapies for breast cancer. Sara?

Thanks, Barb. I'm really excited to be here today. There have been remarkable changes in the management of HER2-positive breast cancer, especially since 2020, with new agents like trastuzumab, deruxtecan and tucatinib demonstrating benefits in pretreated disease. And with the results of DESTINY-Breast09 showing benefits with T-DXd in combination with pertuzumab, we have a new standard of care according to our NCCN guidelines as of this February. What we don't know is how we should be treating patients after they've received ENHERTU. This remains an area of unmet need. On Slide 14, you can see this graph showing the outcomes that we have from patients who've received prior ENHERTU treatment with a median progression-free survival of only 4.1 months. What we don't have are randomized prospective data telling us as clinicians how we should manage patients after T-DXd, but observational and retrospective data sets are fairly consistent, showing that PFS rates of under 6 months is what patients experience post T-DXd. On the next slide, on Slide 15, you can see the list of HER2-directed therapies that we have approved and in late-stage development on the left side. These include HER2-targeted antibodies and bispecifics, HER2-targeted antibody drug conjugates and HER2-targeted tyrosine kinase inhibitors as well. What we don't have are immuno-oncology therapies approved or in late-stage development outside of evorpacept. And so this remains an area that's very exciting to see developed now. On Slide 16, you can see the range of CD47 expression in normal cells versus tumor cells. As explained before, CD47 is expressed on all cell types and is a marker of cells and cancer cells can take advantage of this by overexpressing CD47, allowing them to hide from the immune system. Focusing on the left there, you can see highlighted in blue, breast cancer cell lines, which actually do have higher expression of CD47 in tumors compared to normal, which is highlighted in black. On Slide 17, you can see a meta-analysis of 38 cohorts across 17 publications, which included over 7,000 patients, demonstrating fairly consistently that CD47 overexpression was correlated with shorter survival in patients with cancer. So it's a negative prognostic biomarker. On Slide 18, there are data here relating to CD47 expression in breast cancer, in particular, in HER2-positive breast cancer. On the left, you can see that CD47 expression is higher on HER2-positive breast cancer cells versus HER2-negative on the right. High expression is denoted in the dark blue, moderate expression in green and low expression in blue. So HER2-positive breast cancer cells appear to have a lot more CD47 expressed than HER2-negative breast cancer cells. And in fact, if you look on the right side, you can see that CD47 high cells are more common in recurrent HER2-positive breast cancer shown on the left side compared to primary or initially diagnosed breast cancer. And so these data provide strong rationale for us to be evaluating evorpacept in HER2-positive recurrent or metastatic pretreated disease. On Slide 19, you can see some elegant data demonstrating that T-DXd treatment may actually upregulate expression of CD47, again, providing strong rationale to treat HER2-positive recurrent T-DXd pretreated breast cancer with evorpacept. On Slide 20, you can see the title slide for the data presented recently at ESMO breast. This is an exploratory biomarker analysis from our Phase Ib/II trial of zanidatamab, a HER2-targeted bispecific in combination with evorpacept in patients with HER2-positive metastatic disease. Slide 21 gives you an overview of the study design. This is a Phase Ib/II study, and we presented the safety dose escalation cohort data from 3 patients with HER2-positive breast cancer enrolled and data from Cohort 1, where 21 patients with HER2-positive breast cancer were enrolled who had received at least 3 prior regimens. The exploratory biomarker analyses focused on these 24 patients, and it's notable that all of the HER2-positive cohort received prior T-DXd and a median of 5 prior lines of HER2-targeted therapies. This is extremely heavily pretreated disease. The patients were allowed to enter the study based on prior local HER2 testing of their archival tissue. However, fresh baseline biopsies were obtained in most patients and tested retrospectively for HER2 status by central assessment. And we will be reviewing outcome results based on the central assessment as well as the intent-to-treat population. In addition, patients' tumors were tested for CD47 expression on the fresh baseline tumor biopsies or on archival tissue if fresh biopsies were not available. Patients were treated with zanidatamab at 1,200 milligrams for those patients weighing less than 70 kilograms and at 1,600 milligrams for patients 70 kilograms or greater. The evorpacept was given at 20 mg per kg for these 3 patients enrolled in the Ia or at 30 mg per kg IV every 2 weeks for the rest of the patients. Going on to Slide 22, we're jumping to the efficacy outcome data. What you can see here are the confirmed objective response rate on the top row based on all 24 patients, which is 33%. I find personally these data to be quite impressive in patients who've received a median of 5 prior lines of therapy, including T-DXd. If you go down to the duration of response, those patients who experienced a response had a duration of response of 20 months and a median progression-free survival of 3.6 months. The next column shows the outcomes for the 6 -- excuse me, for the 10 patients who had centrally confirmed HER2-positive breast cancer, where the objective response rate was now 60%. The duration of response was 20.2 months and the median PFS was 8.3 months. These data, I think, are impressive as we reflect back on the fact that our retrospective and observational data suggests that patients pretreated with T-DXd have a median PFS of less than 6 months. So I find these data to be quite compelling and exciting. Those patients who did not have centrally confirmed HER2-positive disease had a lower objective response rate of 14%. On Slide 23, we can now see breakdown based on the tumor CD47 expression levels. Using a cutoff of at least 20% for CD47 expression, there are 5 patients who had both centrally confirmed HER2-positive disease and CD47 positive or greater than or equal to 20% expression. All patients had an objective response in this cohort of 5 patients with a median DOR of 20.2 months. However, those patients whose CD47 expression was less than 20%, that's 4 patients, only one had an objective response. So this is beginning to give us data that CD47 expression level may be a marker to help us select patients particularly responsive to this targeted therapy. On Slide 24, you can see a Kaplan-Meier curve for progression-free survival based on CD47 expression levels. So the progression-free survival with the CD47 expression level of at least 20%, you can see the PFS is 22 months, for those patients less than 20% of the PFS is 3.4 months. Again, although these numbers are very small, we're talking about 9 patients total. The data are quite interesting, and I'm excited to see them hopefully confirmed in the larger study. On Slide 25 are our conclusions for this exploratory analysis of our Phase Ib/II study. Evorpacept and zanidatamab showed promising antitumor activity in patients with heavily pretreated HER2-positive metastatic breast cancer, median of 5 prior lines of therapy, all of whom received prior T-DXd. Greater antitumor activity was seen in patients with centrally confirmed HER2-positive disease. Durable responses were seen in this patient population and seem to be largely observed in patients with higher CD47 expression, supporting a CD47dependent HER2-driven biology that resulted in this prolonged progression-free survival. It's notable that most patients with centrally confirmed HER2-positive disease remained HER2 amplified at progression with T-DXd, supporting the continued use of HER2-targeted therapies. And so a biomarker-driven approach incorporating CD47 may optimize patient selection for this combination regimen and warrants further study. So with that, I will turn it back over to Barb.

Thank you, Sara for that excellent presentation. As we've seen, all of the patients who were centrally assessed as HER2-positive and who overexpressed CD47 in their cancer cells achieved confirmed response to the combination of evorpacept plus zanidatamab. What's new on Slide 27 in the table on the right side of this slide are the individual CD47 expression levels and the confirmed response assessments for each of the 10 patients with centrally assessed HER2-positive disease. An IHC assay was used to quantify the total membrane staining for CD47. A positive result was defined retrospectively as a score of 20% or higher, which we observed in the patients shown in the top 5 rows highlighted with deep blue. All 5 of these patients with CD47 expression responded, including one patient with a complete response. That patient happened to be the only patient from the dose escalation cohort that was centrally assessed as HER2-positive. There was no evidence of CD47 overexpression in the 4 patients highlighted in light gray, one of whom responded to the evorpacept and zanidatamab combination. And then in the last row in this table is a patient who was unevaluable for either response or CD47 expression level. On Slide 28, I want to highlight the consistency that we see between the results of the evorpacept, zanidatamab trial in HER2-positive breast cancer patients and in the previously reported HER2-positive gastric study. On the top half of the slide, you'll see the data that we've just reviewed. This includes a 33% response rate in the 24 patients enrolled based on archival HER2 status, the 60% response rate in the 10 patients centrally assessed as HER2-positive and the 100% response rate in the 5 patients who were also CD47 overexpressing. On the bottom half of this slide, you see the data from ASPEN-06. This is a randomized Phase II study of evorpacept, trastuzumab, paclitaxel and ramucirumab versus the TRP regimen alone. Here, you see the response rate climb as we look across the slide at the same subset. In the ITT, the response rates in patients randomized to the evorpacept arm was 41%. Response rate was over 49% in the patients that retained HER2-positive disease, and it was 65% in patients whose tumors also expressed CD47. And recall, this was a large Phase II study of 127 patients. As you see, the data trends here are consistent across these 2 independent trials of evorpacept when combined with a HER2-targeted antibody. On Slide 29, I'm again showing the results of the Evo-Zani study on the top and the ASPEN-06 HER2-positive gastric study on the bottom. The graph -- these graphs show the duration of response in the subset of patients who were HER2-positive and CD47 overexpressing in the 2 independent studies. The duration of response was remarkable in both settings at 20.2 months and 25.5 months, respectively. On Slide 30, the same consistency across the 2 independent studies as seen with the PFS results in the subset of HER2-positive CD47 high patients. Both studies show a remarkably long PFS, especially in light of the later line settings for both indications. On the left, we see the median PFS for these patients in the evorpacept-zanidatamab study of 22.1 months, while the median PFS for those without CD47 expression was only 3.4 months. And then in the gastric study, shown on the right, the median PFS was 18.4 months for the patients in the evorpacept arm and the hazard ratio in this subset in the randomized trial was 0.39. So these data are really why we're so excited about moving forward with the development of evorpacept in HER2-positive breast cancer. Slide 31 shows our ongoing ASPEN-09 study in the HER2-positive breast cancer setting. Like the zanidatamab-evorpacept study population, all of these patients will have had to have had prior ENHERTU. This is going to -- this is a single-arm study that will enroll up to 120 patients and has a primary endpoint of response rate in the subset of patients who overexpress CD47. We're making very good progress with enrollment, and we're very much on track to achieve our stated milestone of having interim top line data from approximately 80 patients by the middle of 2027. On Slide 32, we show the market opportunity. Ultimately, our aim, of course, is to bring this drug to market. We estimate that there are approximately 20,000 addressable patients who are HER2-positive and overexpress CD47, representing a $2 billion to $4 billion market opportunity in the United States, the 5 major European markets and Japan. I'm now going to turn for the last few minutes of this talk to our second molecule. Slide 34 shows ALX2004, our EGFR antibody drug conjugate. This is actually a very unique molecule. It was designed by protein engineers and cancer biologists within ALX. EGFR is obviously a very validated target, but it's been very difficult to target EGFR in the past with an antibody drug conjugate. We selected amituzumab-derived EGFR antibody to minimize off-tumor skin toxicity and to maximize the therapeutic window. Also, the epitope is distinct from that of FDA-approved EGFR antibodies, minimizing the likelihood of resistance to prior EGFR antibody treatment. The proprietary linker allows for lysosomal cleavage of the linker payload like other deruxtecan ADCs, but the stability of the linker antibody has been enhanced to minimize systemic payload release. The molecule also contains a proprietary TOPO1 inhibitor payload. It has an antibody drug ratio of 8, and the payload has similar direct cytotoxic potency, but it has enhanced bystander activity compared to deruxtecan. On Slide 35, we show the Phase I design. This is an ongoing study, and the study is designed for dose escalation, dose exploration and expansion cohorts. The study is limited to patients who have lung cancer, head and neck cancer, colorectal cancer or esophageal cancer as EGFR expression is higher in these indications. And in this study, we're also making very good progress with enrollment. We are certainly on track to achieve our stated milestone of reporting initial safety data within the second half of 2026. We previously publicly released information that we began dose escalation at 1 milligram. We then said we escalated to 2 milligrams. And then we announced that we had escalated to 4 milligram per kilogram dose cohort. But as we advance the enrollment and dose escalation in this trial, we are now shifting from providing granular updates to reporting initial safety data in the second half of 2026. And I'll now turn the presentation back to Jason.

Thanks, Barb. And also a big thank you to Dr. Hurvitz for joining us today to share her views. At ALX, we're advancing 2 novel cancer treatments with key near-term catalysts that have the potential to be both best-in-class and first-in-class and change care. And with this new data today, we have now demonstrated in 2 studies in 2 HER2-positive cancers that EVO is a very active drug and that a CD47 targeted approach can drive transformational benefit to patients. And again, although not the focus for today, we remain as excited about the potential of our EGFR-targeted ADC, which is also progressing quite well. We now have a stronger balance sheet to fully execute through key milestones over the next year to 2 years, and this really sets us up well for further execution and further catalysts. In sum, Q1 was a strong quarter for us in terms of new data, new leadership within our team and a new financing. So the message for today is one of building momentum as our conviction in both programs remains high, and we're looking forward to additional updates very soon. With that, I'll turn the call back over to the operator for questions. And again, thanks again for the time this morning.

[Operator Instructions] Our first question will come from Michael Yee with UBS Securities.

Congrats on the updates and progress. This is Kaiyue Yang for Michael Yee from UBS. Two for us. The first question for management. So congrats on the data today. So for your upcoming interim data readout in mid-2027 from your ongoing Phase II ASTENBreast study, what would you say is compelling results in terms of ORR and also DOR? And do you plan to put out data on mPFS? And what would be good data for mPFS? The second question for Dr. Hurvitz, could you please help us understand how do you plan to use this drug in the clinical setting if this is approved? Do you expect to test everybody for CD47 expression?

Great. Thanks, Kyle. Appreciate the question. So yes, on the first one, and I'll let Barb weigh in as well. I think we've been very pleased with enrollment to date on our breast study. It's going quite well. Certainly, this data, which we've been able to share just over the last couple of months with the investigators has furthered that enthusiasm. So I think as Barb highlighted well in her comments, feel very good about our time lines and delivering really meaningful data mid-'27. But Barb, do you want to weigh in more on just what we hope in terms of durability and PFS at that time?

Yes, absolutely. Well, as Dr. Hurvitz said, most of the available therapies are not producing very good outcomes for patients after ENHERTU. Much of the data suggests that the response rate may be 15% if you were to receive a trastuzumab chemotherapy treatment alone without evorpacept. So a 30% response rate would be a doubling of what we would expect to see without the evorpacept. So I think anything north of 30% would be a very, very good outcome with a duration of 6 months plus. As you've seen in the data that we reported previously, one of the things that's so compelling about evorpacept is that when we see a response, we see a very good duration of response. By middle of 2027, I think the duration of response will be pretty preliminary. So I'm expecting that with 80 patients worth of data that we can assess response rate pretty well. Duration of response may be somewhat truncated, but I would hope to see somewhere in the range of 6 months plus with immature data. And I would think that PFS might be too immature to report, but it really depends on the continued progress. So we'll have to see. Yes. And Dr. Hurvitz, whether you have comments on your expectations for the ASPEN-09 data or in particular, the second part of the question about the role of CD47 testing in the clinical practice if evorpacept is able to get to market.

Yes. Thank you, Barb. I fully agree with your -- what you stated about the ASPEN-09 study. I mean I think seeing anything around 30% objective response rate and a PFS of 6 months would be pretty important in patients that heavily pretreated. So I echo your comments. In terms of the additional question regarding how we would use CD47 expression, I think this is something that's really very compelling and interesting. And the ASPEN-09 study is going to allow patients, as Barb stated, with both CD47 low and high tumor expression so that we can validate these data. We have to keep in mind, it was a small number of patients, but the data were really pretty interesting and compelling. So if CD47 is validated as being a predictive marker for benefit from evorpacept, then I do think we will be probably seeing it developed as a companion diagnostic. But it's too little data at this point to see -- to make that statement definitively that that's how we'll be doing -- using it. I've been treating breast cancer patients for over 20 years. And the data with ENHERTU were really exciting when we first saw them come out in 2019. I feel that same flurry of excitement seeing these data because it's pretty tough to treat patients post T-DXd now. Their progression-free survival, their duration of response to the next line of therapy is pretty short. And so these data provide patients a lot of hope. So I'm excited to see the data validated just taking into consideration that it is a small number of patients.

Our next question will come from Derek Archila with Wells Fargo.

Congrats on the update here. Just a few questions from us. I think, first, you alluded in the prepared remarks, but just remind us the population in ASPEN-09 and then how it compares to some of the patients from the update today, that would be helpful. And then just 2 quick ones. So I think you mentioned the 20,000 patients CD47 high post HER2. So does that assume greater than 20% CD47 expression? Or I just want to make sure the cutoff is the same? And then with the data here and the proof of concept with the zani combo, I guess, how do you think about further developing this?

Yes, sure. Those are great questions. Thanks, Derek. So I think in terms of the population that we're studying now, it is this population. That's what I think really derisk, if you will, this current study. I think as we walk through today, this is a late-line breast population. 100% of the patients in this cohort had seen prior in HER2. On average, it's seen 5 prior lines of HER2-directed therapy, et cetera. So I think the beauty of this design that we're pursuing now is it's effectively the same. So I do think that definitely contributes to our confidence in this study. Then on the second question, and I can have Barb weigh in on this. I think we do know from the literature, we expect that the expression profiles will be different across different tumor types. That's certainly what we've seen when we look at breast and gastric. But again, I think what's so encouraging is that there's just a lot of room here in terms of choosing a cutoff, if you will. So Barb, do you want to speak to that a little more?

Yes. I think one of the most important things about what we will understand with the data from ASPEN-09 is what's the optimal cut point? Where on the total membrane staining scale, would we see the greatest differential between high response rate in the CD47 positive versus CD47 low. And it's likely to be a continuum, and we're likely to be able to not see a cliff where it's crystal clear. What we've seen in both the gastric and the breast study, though, is we're seeing about half of patients overexpress CD47. At least in our gastric study with a larger sample size, we shared at SITC in the fall of 2025, a variety of different cut points that range from 45% to 57%. Gastric may be different than breast, though. And so with the Zani data, we have just such a small number of patients. So -- but what we saw in the Zani trial, again, about 50% were CD47 high. One of the biggest reasons we increased the sample size from 80 to 100 is just so that we could get more data because this point exactly what is the cut point? How does it differentiate response rates? And what does it do to the population in terms of potential commercial opportunity balanced against finding the best treatment effect for patients is a really important question. So ASPEN-09 will help us optimize the cut point. My best guess today is it will be somewhere in the range of 50%, 60%, 70% of patients could be CD47 overexpressing, but we'll see. What also, of course, are the 2 slides that Dr. Hurvitz showed where the rates are higher in patients who have more advanced disease. They are higher in the cell lines post ENHERTU, and they are higher in patients who are HER2 positive. So it could even be higher than 50%. We'll learn a lot from the ASPEN-09 on that exact question.

And I think your last question, Derek, just on zani, a couple of comments there. I think, one, really pleased with the partnership with Jazz and what we're seeing here for patients. I think no question that very strong activity with this combination. The second thing, and I think why it's important to look back at our mechanism is EVO works in combination with an Fc active antibody, and it was designed to do so. And so I think what's really promising about the totality of the data is whether we're combining EVO with traz, with zani, with Rituxan, with an anti-CD38 like Sarclisa, we see activity. And I think that's right on mechanism. And so for us, taking the combination of Evo plus traz going forward makes sense. Again, it's the exact same combination that we utilized in our ASPEN-6 study, which is a randomized study, as you know. So I do think that gives us a lot of conviction here going forward.

Moving next to Allison Bratzel with Piper Sandler.

This is Ashleigh on for Ally. Just two questions from us. So just curious to learn what's the latest on the companion diagnostic development. I know you guys have talked about this on previous earnings calls. So just looking for some updated thoughts there. And also just based on the data today and this data set, it looks like the CD47 biomarker strategy is being further validated. So does this impact the statistical powering at all for ASPEN-09?

Sure. That's great question. Do you want to go on the CDx front, Barb?

Yes. Sorry, Jason. Yes, the Ventana Laboratory is our partner. We are working right now on developing a companion diagnostic. What we want is by the time we are finishing the full data set of ASPEN-09, as I alluded to earlier, setting a cut point that could then be available for preselection of patients potentially in a Phase III trial. That's our base case assumption is that our Phase III would be in preselected patients. So we will have an assay available. And then if that Phase III study is positive, of course, then we would work with our Ventana partner to not only bring evorpacept to market, but to ensure that the companion diagnostic were then commercially available. In terms of statistical powering -- go ahead. Go ahead, Jason.

No, go ahead, Barb. I just -- I mean, to kick that off, I mean, I think, of course, we're not going to power the study at 100% ORR, and that's not what we're going to do. And I think -- but if you look at the beauty of a targeted approach, I think it allows and our hope is it will allow a much more targeted Phase III where we can run a tighter registrational study with assumptions that I think are reflective of what we've seen, which has been a really strong spread, if you will, both in terms of ORR as well as DOR and PFS. But go ahead, Barb, I know I cut you off there.

No, no, that's exactly right. I think with a single-arm trial, ASPEN-09, the statistics are really around estimating the point estimate of the response rate and narrowing the confidence intervals with a larger sample size. But what ASPEN-09 does is it allows us potentially greater precision with this larger trial of 120 patients such that we can then have a lot of confidence walking into a Phase III subsequently and really understanding the treatment effect in patients based on their CD47 status.

Moving next to Li Watsek with Cantor.

This is Daniel Bronder on for Li Watsek. We have a couple, one more technical and then one regulatory. We'll start with the technical one. On the poster from ESMO breast, you have the concordance of HER2 amplification by ctDNA and FISH. I was just curious how that translates into HER2 positivity in IHC. And then on the regulatory front, you mentioned registrational trials a few times in your opening remarks and in the previous response as well. Does that mean you're ruling out a potential accelerated approval based on ASPEN-09? Or are you keeping optionality here?

Those are great questions. Go ahead.

I'm sorry, I should stop doing that. And Dr. Hurvitz, I'll have you make some comments after I'll address the regulatory one first, give you a chance, but I'll have you talk a little bit to the HER2 data that was in the poster. So the regulatory strategy, Jason said a few minutes ago that we do not have an expectation that we could replicate 100% response rate. That would be fabulous if we could. And if we do, I think absolutely, we would talk to the health authorities, U.S. FDA about an accelerated approval strategy. I just think that, that is such a high bar, and we're also very, very aware that an advantage for patients would be anything north of, say, 30%. So we'll just have to wait and see what those response rate -- what the response rate is in ASPEN-09. I think regulatory approvals for single-arm trials when you have a combination of two agents, very difficult to get an accelerated approval. It's also true that it's very difficult to get an accelerated approval with retrospective application of a companion diagnostic. We would need at least some additional prospectively selected patients. So there are a lot of regulatory considerations. But I think at this case, that is not our base case. Our base case is that a Phase III trial will be required because the bar is high for accelerated approval, but we'll just have to wait and see what the response rate is. If it's extraordinary, we'll talk about it. If it's somewhere north of 30%, but less than 100%, it's going to be an uphill battle. But obviously, that would benefit patients if we could bring it to market sooner, but we'll wait and see. So maybe I'll flip it back to Dr. Hurvitz to talk a little bit about some of this HER2 data in the poster.

Yes. Thank you, Barb. The HER2 data is actually kind of interesting. So there are 2 aspects. First, there was a difference in the central confirmation of HER2-positive disease that was shown in the poster where 10 of the patients out of the 24 were noted to have centrally confirmed HER2-positive disease and 14 were not. And I would just call out that this is not uncommon. It's been reported in other studies that there is discordance. And it may not necessarily mean that there's been HER2 loss that's occurred over time. It may actually be related to artifact from testing older tissue. It's important to keep in mind that in order for patients to enroll in the study, HER2 status was based on archival tissue that was read locally. But once they got on the study, the assays that were done were done on fresh biopsy samples obtained at the start of the study from most patients, and those were tested centrally. And so the testing changed from local to central and also from archival tissue to fresh tissue after in HER2 for the vast majority of subjects. So either of those factors could have contributed to the differences between local HER2 positive results and central positive results. Another point was in the poster in Table 2, there were some data presented looking at concordance between HER2 amplification status by the central tissue-based FISH as well as plasma ctDNA next-generation sequencing. And it's important to call out that of the 6 patients with HER2 amplification by FISH -- or excuse me, the 7 patients, 6 of them were also noted to have amplification by ctDNA. Just one of the patients there did not, and that may have just been due to the fact that there wasn't enough circulating. There was a low tumor fraction. It was below the threshold of expression. But those that were negative centrally by FISH were also negative by ctDNA. And so that's reassuring and provides sort of validation that the central HER2 amplification is correlating well with circulating tumor DNA.

We'll go to our next question from Roger Song with Jefferies.

Congrats on the positive data and excited about the momentum going into ASPEN-09. Maybe two from us. So you previously mentioned we're seeing roughly half of HER2-positive patients being CD47 high. As we're into the early ASPEN-09 screening, is the observations you're seeing consistent with that half of HER2-positive patients being CD47 high? And then is tissue adequacy and turnaround time any potential gating factor here for enrollment?

Yes, sure. Good question. So I think on the ASPEN-09 front, of course, we're going to monitor that closely. And then in terms of the disclosure, we'll wait and disclose that when we get to the data readout on 80 patients. Obviously, I think it's something that's important, and we're able to monitor really quite quickly. So the lag has not been a concern. And of course, I think our team is focused on that. But Barb, do you want to add anything there?

Only that the ASPEN-09 study enrolls patients irrespective of CD47. So there is no urgency in terms of the turnaround. You're absolutely right that it would be something that we need to ensure that we work through those processes for a future Phase III, assuming that, that future Phase III requires CD47 overexpression for entry. But this study does not. So the urgency is not there. But as Jason says, we're not going to disclose prevalence rates or any of the biomarker data until we come out with data in the middle of 2027.

Moving next to Sam Slutsky with LifeSci Capital.

This is Oliver McCammon on for Sam Slutsky. So in the Phase Ib/II breast cancer study, you defined CD47 high as total membrane staining of 20% or greater. And in the past, you talked about 10% or greater IHC 3+ as being a cutoff for CD47 high. Can you just discuss the nuances of the 2 cutoffs used and the selection process?

Sure. Thanks, Oliver. Again, I think Barb hit on this, but we fully anticipated even ahead of this data that the expression profiles would be different. There's, of course, a lot of similarities between HER2 positive tumor types. Insofar as typically a drug that works in one will work in the other. That said, I think there will be a difference in expression, and we're seeing that here. And again, I would just highlight that if you think about defining a cutoff when you see ORR as high as we're seeing across both of those studies, it suggests there's probably many different right answers, if you will, in terms of the cutoff we could use. But Barb, do you want to expand on that?

Just in terms of total membrane staining and how that's calculated. So the proportion of cells that are 1 plus are multiplied by 1, the proportion of cells that are 2 plus IHC stain staining intensity would be multiplied by 2 and so forth. So IHC 3, multiply the proportion of cells that are IHC 3+ by -- and then you add those. So it's a cumulative total. So if I were to look at the threshold for gastric 10% or more of cells being 3 plus, that would potentially be about a score of 30. It may be that different labs and slightly different methods. But to put it in context, it's similar, but it's -- cut point was a little bit lower in the zanidatamab trial. It was zanidatamab-evorpacept was a very small trial. So we will look again once we have data from ASPEN-09 and optimize that. So it may change. In fact, that likely change much, much sample size, but technically we are doing it. And we modified it a bit once we start work towards development [indiscernible] with our partners at [indiscernible].

This now concludes our question-and-answer session. I would like to turn the floor back over to Jason Lettmann for closing comments.

Great. Thanks, everybody. It was a strong quarter for us. I appreciate all the engagement here today. Both programs, as we mentioned, are on track and super excited about what we're seeing in both. So again, I appreciate the questions and support and looking forward to future updates. Enjoy your Friday and your weekend. Thanks so much.

Ladies and gentlemen, thank you for your participation. This does conclude today's teleconference. You may disconnect your lines, and have a wonderful day.