Alx Oncology Holdings Inc Q2 FY2025 Earnings Call

Thank you all for joining us today to discuss our Q2 2025 results. We're looking forward to sharing our updates with you this afternoon. Before we begin, please take a moment to review our forward-looking statements. Our agenda for today will include an update on our key accomplishments in the second quarter of 2025. We are particularly pleased to present new data from our ASPEN-06 trial, which highlights CD47 expression as an important predictive biomarker for enhanced clinical response to evorpacept. Our main objectives today are to share this new data, discuss its implications for our development strategy for Evo, provide an update on our novel EGFR-targeted ADC, ALX2004, which is about to enter clinical trials, and offer revised guidance on our financials and cash position. Alan, our CMO, will kick off with a presentation on these top-line results and clinical program updates related to Evo. Harish will then follow up with an update on our cash runway and key milestones. Finally, we will open the floor for your questions. In the second quarter, we made significant progress with our evorpacept and ALX2004 clinical programs. We are excited to share data indicating the potential of CD47 expression as a predictive biomarker and the opportunity to pinpoint patients who will benefit the most from evorpacept. Alan will provide detailed insights in his section, but our analysis shows that patients with high CD47 expression experience the most significant benefits from evorpacept compared to those with lower expression. Consequently, we have adjusted our Phase II clinical trial in breast cancer to focus on a CD47 and HER2 biomarker-driven approach. These findings pave the way for targeted oncology strategies with Evo in various tumor types, especially considering the widespread overexpression of CD47 in solid tumors and hematologic malignancies, which positions Evo as a promising targeted IO therapy. Regarding our progress, we are on schedule to dose our first patient with ALX2004, a unique EGFR antibody ADC, and we are eager to initiate clinical trials this month. Additionally, we are pleased with the progress in our partnered randomized Phase I/II UMBRELLA study with Sanofi, which is assessing evorpacept alongside SARCLISA and dexamethasone in patients with previously treated multiple myeloma. The initial phase of this study is now complete, and Sanofi will be moving into the dose optimization stage with that combination. On the financial side, our focus on evorpacept in breast cancer and ALX2004 has allowed us to extend our cash runway guidance into the first quarter of 2027, strategically positioning us to achieve several key data milestones later this year and next year. Turning to Slide 5, I want to highlight our findings from the CD47 expression analysis. We now have strong evidence that CD47 overexpression serves as a reliable predictive biomarker for patients. Earlier this year in January, we presented the top-line results for ASPEN-06 at ASCO GI, and today we are revealing new data from a preplanned analysis of that dataset. In this analysis, we assessed CD47 expression levels using IHC and discovered that patients with confirmed HER2 positivity and high CD47 expression exhibited a significant response to evorpacept compared to the control group. The overall response rate for the intention-to-treat group was 41% compared to 27%. For patients with clearly high CD47 expression, the benefit was even more pronounced, with an overall response rate of 65% versus 26% in the control group, and a nominal p-value of less than 0.05. This substantial benefit for evorpacept, in combination with TRP for CD47-high patients, was also observed in terms of duration of response, progression-free survival, and overall survival. We plan to present the comprehensive dataset at an upcoming medical conference in the fourth quarter of this year. Turning to Slide 6. As a reminder, Evo is a unique CD47 blocker that sets it apart from CD47 inhibitors that have been developed in the past with its differentiated safety profile and clinical activity. With its inactive Fc domain, which delivers best-in-class safety and combination potential across a wide range of biologics and targets, it's truly a different CD47, and we're now seeing that in the clinic. Following the encouraging results that we have shared today and are looking forward to sharing more later this year, we clearly see that CD47 is a biomarker. This biomarker-driven enhanced efficacy from Evo in HER2-positive gastric is a clear demonstration of the drug's potential to be a first-in-class and best-in-class targeted IO therapeutic and to drive superior outcomes for patients with CD47 overexpressing cancers. Turning to Slide 7. CD47 is overexpressed across a wide range of both solid and liquid tumors. This has been well published in the literature, and it's clear that CD47 is a marker that both solid and liquid tumors use to evade the immune system. As you can see in this slide, CD47 is overexpressed across a very wide range of solid and hematologic malignancies, making it a very compelling target, not only where we're focused but beyond. Now turning to Slide 8. It is also clear that this overexpression matters. When you look at research in CD47 over the last decade plus, it's a very strong foundation that CD47 is a negative prognostic biomarker. What you can see here is in the meta-analysis of 38 cohorts across 17 publications, which includes over 7,000 patients, CD47 is clearly associated with shorter survival and worse outcomes. You can see on the right the wide range of tumor types where this has been documented. So, turning to Slide 9. This sums up what we're most excited about now, which is really driving a targeted IO breakthrough and a first-in-class drug with Evo.

Thank you, Jason. Good afternoon, everyone. I'm excited to be able to share with you some updated data from ASPEN-06 on CD47 expression as a predictive biomarker for evorpacept. We'll go to the next slide. This slide is a quick highlight again to review the evorpacept's mechanism of action with anticancer antibodies and specifically with HERCEPTIN on this slide. Again, the mechanism is relatively simple and the concept being that due to CD47 expression in healthy cells driving on-target toxicity, the conventional approaches with an active Fc were fraught with significant challenges related to toxicity. As a result, evorpacept was designed to have a silent Fc and utilize the companion cancer antigen-directed antibody as a means to direct the activated macrophages to the cancer target. Next slide. Again, what I'd like to show you today is to go back over some of the data that we have with respect to our ASPEN-06 data and will highlight the impact of CD47 expression as a key predictive biomarker to enhance the activity of evorpacept we previously discussed in the gastric GE junction space. This slide reminds you all that patients were eligible based on the archival tissue for HER2 levels of expression, and there were 127 patients on the study. Next slide. To remind you, the addition of evorpacept drove a 41% response rate as compared to 27% in the control arm. Additionally, as noted in the swim lanes, both qualitatively and quantitatively, there was an improvement in the duration of response. The median going from 9.1 months in the control arm to 15.7 months in the experimental arm. Now, the next slide and an important point here is, these are now patients that had a confirmed HER2 positivity as noted by fresh biopsy or ctDNA. This resulted in a further increase in the response rate to 49% on the experimental arm as compared to 24.5% in the control arm and still maintaining an excellent duration of response of 9.1 months in the control and 15.7 months in the experimental arm. I'll also comment on the fact that you see despite the fact of confirmed HER2 expression, there was no difference in response rate or DOR on the control arm. Moving to Slide 15. Patients in ASPEN-06, as noted previously, were tested for both HER2 and for CD47 expression. Let me now show you how we got there. The ITT population are patients enrolled with either HER2 fresh or archival tissue and it totaled 127 patients. We then attempted to confirm those patients that were truly HER2 positive closest to starting on study. These patients were defined by a fresh biopsy or by ctDNA that was collected at the time of entrance into study. This resulted in 96 patients confirmed to be HER2 positive. Of those 96 patients, 90 patients had tissue evaluable for CD47 assessment by IHC. 43 of those patients were defined as CD47-high, as defined by greater than or equal to 10% of the cells being IHC 3+. 47 patients were CD47 low. So now, let's take a look at the results on Slide 16 for CD47 expression as a predictive biomarker. You'll note in the ITT population, as a reminder, the response rate was 41% versus 27%. Now when we combine CD47-high along with the confirmed HER2 positivity, the response rate is up to 65%, and you'll note no difference in the control arm with a response rate of around 26%. We'll also note that DOR, PFS and OS showed a strong magnitude of benefit as well in this population. The results were also consistent across multiple CD47 expression cutoffs and the full data set will be presented at an upcoming medical conference in the fourth quarter of this year. Moving to the next slide. The magnitude of benefit that we've seen in patients with high CD47 expression in HER2-positive gastric cancer enables us to pursue a targeted development strategy for evorpacept in breast cancer and other tumors that overexpress CD47. Moving to the next slide. So, I'll now discuss our clinical program for evorpacept in breast cancer, emphasizing the CD47 biomarker strategy. Our opportunity in breast cancer now has a high probability of success, having been derisked by positive data in 2 different HER2-positive cancers. In addition, this represents a very high unmet need as the changing frontline standard of care drives opportunity in those patients who have progressed on ENHERTU and/or other HER2-directed therapies. In addition, it remains a highly targeted approach now using both HER2 as well as CD47 in a biomarker-driven approach. Next slide is Slide 20, where we'll discuss our prior data in zanidatamab in combination with evorpacept in patients with metastatic breast cancer progressing on prior HER2-directed therapy. The key eligibility criteria are shown on the left, and it required at least 3 prior regimens, including HER2-directed therapy. There were 21 patients who are HER2 positive in this clinical trial, which we'll emphasize, and I'll discuss in the next slide or 2. Now looking at these 21 patients, if you look at the overall response rate, it was 33% or 7 out of 21. But importantly, given the mechanism of evorpacept and requiring HER2 expression, you'll note in the 9 patients who were confirmed to be HER2 positive, there were responses in 5 of those patients for a 56% response rate. The median DOR was not reached and the median PFS was 7.4 months. This compares favorably to the Phase III SOPHIA study that looked at margetuximab in single-agent chemotherapy or trastuzumab and single-agent chemotherapy. Now, let's discuss the importance of CD47 with respect to HER2-positive breast cancer. Slide 22 shows that several studies have now shown that CD47 protein expression in HER2-positive breast cancer is overexpressed at the time of initial diagnosis, somewhere in the range of slightly over 50%. Slide 23 looks at CD47 and it's being upregulated in response to prior ENHERTU therapy in this HER2-positive setting, looking at breast cancer cell lines shown here. This is, of course, important for our upcoming breast cancer study that is going to be targeting post ENHERTU patients. Slide 24 has 2 key elements of importance. First, CD47 expression is higher in HER2-positive breast cancer cells as compared to HER2-negative, shown on the left. In addition, CD47-high cells are more common in recurrent and previously treated HER2-positive breast cancer. Slide 25 is the schema of our proposed breast cancer study. Based on the magnitude of benefit for evorpacept in CD47-high patients in ASPEN-06, we have amended the design of our Phase II ASPEN breast cancer study in HER2-positive patients now evaluating evorpacept in combination with trastuzumab and single-agent chemotherapy. This updated trial will now be a single-arm design, enrolling HER2-positive patients by archival biopsy and will be evaluated with CD47 expression. With this new design, we expect to accelerate enrollment into the study and provide interim data in Q3 of next year. Evaluation of CD47 expression or evorpacept's benefit is expected to provide results that could support a biomarker-driven registrational study in this indication.

Thanks, Alan. Now turning to the commercial impacts and the breakdown of the market in terms of CD47-high in HER2-positive breast patients. As you can see in the slide, in the big bubble, there are roughly 48,000 patients in the second line plus setting who are HER2 positive. Of that, there are approximately 60% to 80% of patients that retain HER2 positivity. And of that, there are 50% to 70% of those patients are CD47-high. As Alan highlighted, there were a number of publications to support CD47 overexpression in HER2-positive breast. We believe this represents roughly 20,000 addressable patients who are both HER2-positive as well as CD47-high. If you boil this down and use conventional estimates on pricing, you get to roughly a $2 billion to $4 billion market opportunity, again, just in patients that are CD47-high and HER2 positive. This represents a significant opportunity for Evo. This is one of the aspects that we're excited about as we think about Evo as a targeted IO approach. So, turning to the next few slides, I wanted to quickly just give a quick update on ALX2004. We're very excited about advancing our second pipeline program here. As you remember, it is a novel ADC that is targeting EGFR. On Slide 28, highlights the development approach and how we got here. As you may recall, ALX2004 was developed by a world-class team in our Palo Alto labs with a vision starting back in 2021 to create a best and potentially first-in-class drug designed to maximize the therapeutic window and overcome the historical tox challenges that others have encountered in targeting EGFR with an ADC. ALX2004 has been optimized on all 3 components to do this; the payload, the linker, the antibody have all been optimized to create a truly novel molecule against a validated target. On the antibody front, its affinity was tuned to maximize the therapeutic window with the EGFR binding epitope that is distinct from the approved EGFR antibodies. The linker payload construct was also optimized to be similar to ENHERTU in many ways yet offer enhanced bystander effect with improved linker stability for on-target delivery of payload. This then translated to preclinical experiments, supporting dose-dependent activity and a differentiated safety profile that supports our conviction that this molecule could potentially demonstrate efficacy with, importantly, a manageable safety profile in patients. So, turning to Slide 29 quickly. This, again, highlights our development plan. Again, we are on track and continue to execute on our timelines that we communicated before as we are very close to dosing our first patient and expect that to happen later this month. We will then go through dose escalation, dosage exploration, and ultimately, dose expansion, with a goal of really nailing the dose here that is safe as well as demonstrating efficacy across 4 different tumor types that are known to be EGFR expressing. This will then set up the program well to advance into a registrational study after this. So, that's the plan with ALX2004. Again, very excited that we remain on track and looking forward to dosing our first patient very soon. I will now turn the call over to our CFO, Harish, who will walk through upcoming milestones as well as revised guidance on our financials.

Thank you, Jason. Now, turning to Page 31. Here's a snapshot of our clinical pipeline. As we have communicated previously, we are pursuing a focused development strategy for evorpacept in combination with anticancer antibodies, given the consistent proof of concept that we have seen in various clinical studies with different monoclonal antibodies. Our priority is to advance our HER2-positive breast cancer program with a CD47 biomarker-driven approach. From our partner program with Sanofi's SARCLISA in multiple myeloma, dose escalation has now been completed, and we are excited to see the trial moving into dose optimization phase. ALX2004, our EGFR ADC program, remains on track to dose the first patient this month. Now, turning to the next slide, 32. You can see the major data readouts from our clinical pipeline and potential inflection points we have over the next year. As we mentioned multiple times during this call, we are excited about the CD47 biomarker data from ASPEN-06, and we'll be initiating our amended breast cancer trial, which will include a CD47 expression analysis. We expect to start dosing patients in Q4 this year and expect interim data readout from this trial in Q3 2026. On the ALX2004 front, we're about to dose our first patient this month, as I mentioned, and anticipate providing initial safety data in the first half of next year. With the prioritization of our capital to focus only on evorpacept trial in breast cancer and on ALX2004 and halting our previously announced evorpacept colon cancer program, we have been able to extend our cash runway guidance into the first quarter of 2027. Please note that the cash runway guidance here reflects a significant decline in clinical trial spend in future quarters as we close out multiple legacy evorpacept trials, including ASPEN-03, -04, ASPEN-06 and ASPEN-07 following the final data readout in these trials. You can refer to our Q2 '25 results press release for a review of our detailed financial statements for the quarter. With this, I'd like to turn the call back to Jason for any closing comments before opening the line for Q&A.

Thanks, Harish. Appreciate it. In summary, again, really strong quarter in terms of our execution, continued tight discipline around our capital, and I think a number of milestones that we've achieved. Of course, great to have validation with Sanofi and our study with SARCLISA, excited about next steps there, continue to execute on ALX2004, and as we mentioned, very close to FPI. And then last, as Harish mentioned, continue to be very mindful of cash and pleased to have now extended our cash runway to Q1 of '27. Overall, though, I'd say we're most excited about what we're seeing with CD47, again, as a predictive biomarker. As many of you know, IO has been a tough space for some as it's been difficult to target the right patients and to know where and how to best deliver those drugs. What we see here with Evo and with CD47-high is, we know we have a biomarker that negatively impacts patients. To be able to select those patients, treat them with Evo, again, very much on mechanism when we think about CD47, really offers us the potential to deliver a transformational benefit in our study in breast. So, excited about that, looking forward to presenting the data in Q4 at a medical meeting, and we're looking forward to any questions from you all. So, with that, I'll open up the floor to Q&A. Thank you.

Our first question comes from Allison Bratzel with Piper Sandler.

Really just a couple of clarifications. First, do you need to meet with the FDA to finalize trial design changes for ASPEN-Breast? Or are there any other barriers to implementing the biomarker strategy in the trial? Secondly, could you just frame for us investigator feedback on the CD47 biomarker approach? How you think that it could affect enrollment trends? And then lastly, just on the interim look for ASPEN-Breast in Q3 next year. Maybe just frame for us what kind of data you would need to see to give you confidence to continue pursuing that biomarker approach in breast or other tumor types?

Great. Thank you. I appreciate the great questions, Ally. So, on the first one in terms of FDA feedback and interaction, we did submit the amendment for the ASPEN-Breast study to the FDA and are ready to go there. It was important for us to revise the protocol, given this new data, to make sure we're really focused on CD47 as a biomarker there. There's no holdup from the FDA, and we're all set to move forward, which I think is good news. So then, on the second question on investigator feedback, I'd say it's been great. Even before this, there was a lot of enthusiasm for Evo and for CD47. Again, there's a good reason why CD47 was such a hot target for so long, because there's a lot of fundamental biology here. When you think about the future of breast cancer, understanding why patients progress on ENHERTU and identifying evasion modes is really important for clinicians and, of course, for patients. What this represents for patients that progress on ENHERTU is a big open question as to what's next. For us to be able to deliver this as a biomarker, I think, is really exciting. We've had phenomenal feedback. We are cautious about how that will translate to enrollment. But again, I think it's a unique story we're going to be able to tell. There is competition, a lot of it is just other ways to go after HER2 with ADCs. I do think due to the competitive framework and shakeout here, we're pretty much the only answer for patients that are CD47-high. So, I think the clinician feedback has been great. On the last, the interim, I'll let Alan weigh in on the bar and what we're hoping to see. But we're laser-focused on this study from an execution perspective, going to enroll it as quickly as we can and deliver data Q3 that's meaningful. But Alan, do you want to talk more about specifics around what we're hoping to see?

Yes, sure. Thanks, Jason, and thanks for the question. Historically, the data that is generally accepted to be the current standard of care with HERCEPTIN and single-agent chemotherapy is roughly around 20%, give or take 1% or 2% on either side. What we're looking for is something that we think would be fundamentally changing. That would probably be something in at least the high 30s and preferably 40% or greater, which would give us an opportunity that we think would be both statistically noteworthy and certainly clinically compelling as well. We would anticipate those response rates to be associated with durability as well.

Great. Did that answer the questions? I think I got it.

Our next question comes from the line of Li Watsek from Cantor.

Thanks for sharing the new data. I have a couple on the ASPEN-Breast study as well. Can you clarify if you intend to pursue a registrational path with the current Phase II study or will you need to run a separate study to support registration? And then, how do you address the contribution of components question for the study?

Sure. Thanks, Li. Those are both great questions. I'll let Alan offer some thoughts on whether or not this could be registrational. I think it really would require an accelerated approval path, but we can talk more about that. At a high level, the goal of this study is really for decision-making purposes around everything we've been discussing. We have very exciting data with CD47, and importantly, this wasn't cherry-picking in that we saw the same effect across multiple cutoffs of expression. The data is robust. But of course, to go into Phase III, you need to determine your cutoff. This will be informative for us to be able to do that and then run a biomarker-driven registrational study. But maybe I'll let Alan add some color on the accelerated approval potential here.

Right. Sure. It all depends on the data. As I mentioned, looking at response rates that are in the 40% or higher range would certainly leave a door open for discussion. However, the study that we have currently is for 80 patients and using a percentage of those would be both HER2 positive and CD47-high. As we look at the data, one of the advantages of this single-arm study is there's no defined interim analysis. It's an open-label study. If we're looking at data that seems compelling, we could add patients so that it might have an opportunity for accelerated approval. But that, of course, will depend on the data we've mentioned. Your other question relates to contribution of components. The study is HERCEPTIN and single-agent chemotherapy of dealer's choice. There are about 4 or 5 choices that the investigator can have. The difference would be Evo over what has been utilized as standard of care. We think that the contribution of components would be fairly well established with this study. Did I address both of your questions?

Yes.

Our next question comes from the line of Roger Song from Jefferies.

Congrats for the data. I just have a question related to the CD47 cutoff, 10%. Since you mentioned you looked at different expression levels, how do those data look? Would you expect this 10% cutoff to apply to other Evo programs, including different combos and different lines of therapy? Lastly, any biology reason for higher or lower expression level as the cutoff?

Thank you, Roger, for the great questions. It's crucial for us to consider various cutoffs since there's no clear consensus on what constitutes sufficient levels. We'll present more information at a medical meeting in Q4. The key point Alan mentioned is that across multiple cutoffs, our findings are consistent. We've shared data on overall response rate, but importantly, we're also observing promising results for duration of response, progression-free survival, and overall survival, even when our power to demonstrate benefits is limited, particularly for PFS and OS. We could have selected different cutoffs, and the results would likely have been similar. As for which cutoff we will use for other studies, that is yet to be decided. However, it appears that the choice of cutoff may not significantly impact our findings. Regarding the biology aspect, Alan provided an overview, and it's beneficial to target something well-studied, like CD47. The five studies we've presented raise important questions about CD47 expression in breast cancer. We believe this evidence is quite strong, and our upcoming breast study should provide further clarity on this issue.

Yes, it does.

Our next question comes from the line of Sam Slutsky from LifeSci Capital.

I guess first, did you look at this correlation between CD47 expression and response in any of the other studies with evorpacept? If so, did the data replicate? If I missed this, how many patients do you expect in the interim with ASPEN-Breast?

Yes, sure. We can take the second one first. I think what we're hoping to do is show it gets to meaningful data. I don't think it's a handful of patients. I don't think it's going to be the full data set either. If you look at how many patients we shared here, we're seeing a strong signal in a relatively small data set. This is not a formal predefined interim analysis. This will be driven by the data itself. We are actively looking at this question of how CD47 plays in our other data sets. As you know, Sam, we've been testing different mechanisms here; one with an antibody, two with a checkpoint, three with an ADC. We need to be cautious about drawing too many conclusions across different mechanisms. Certainly, when we think about the work with Sanofi or Jazz, looking at CD47 there is very important. More to come, Alan, anything on that last one that I missed or that you want to add?

No, Jason, I think you covered it well. Sam, have we addressed your questions?

Yes. So, might we see that then at a future update if any other studies saw any correlation?

Yes. No, absolutely. It's an important question and something we're actively on top of. Great thought.

Our next question comes from the line of Arthur He with H.C. Wainwright.

This is Arthur on for RK. Obviously, very encouraging and intriguing data point for the CD47-high. I'm just curious, one question regarding the data analysis. How is the baseline characteristic between the CD47-high and low patient group?

So, the baseline characteristics, we'll share more at a medical meeting as well. I'd say they're generally well balanced. When we cut this analysis a number of different ways, I think it stands up. If you think about expression levels and different cutoffs, as you widen the aperture, so to speak, you're going to include more patients. When we do that, again, looking at different cutoffs is important because it's robust. We'll have more to say on the baseline front, but at a medical meeting, we didn't disclose that here. But again, I think the data holds up.

Got you. My second question is regarding Slide 23. The upregulation of CD47 expression following ENHERTU treatment is very interesting. Do you have any information or know of any patient biopsy data concerning the CD47 expression after the treatment?

So, I think your question is about patient biopsy data post ENHERTU?

Yes. There is evidence for that, not from our studies, but there is evidence that the CD47 levels of expression go higher with subsequent lines of therapy, potentially as a means of resistance.

Yes. That bodes well for the clinical trial that we're going to be conducting post ENHERTU and in previously treated patients. We're going to also encourage repeat biopsies when we can. Of course, that can be a pain for patients and clinicians. But if we can get data both on biopsy at the time of diagnosis and when they enter the study, I think that will be a really helpful tool. The 50% number across the 5 studies is a fair assumption. If you assume that CD47 is, in fact, a resistance mechanism and it's actually higher; then, in the overall population, we could see as high as 70% of the patients, but that's to be determined.

Our next question comes from the line of Ting Liu with UBS.

I have a quick follow-up on the 10% IHC3 cutoff. It sounded like this biomarker analysis was pre-specified, but this threshold wasn't. If you could clarify quickly, am I understanding this correctly? And when you make the protocol change to the breast cancer study, would you pre-specify a threshold? Or will that be dependent on the data analysis?

Yes. No, that's great. Those are great questions. Thanks. It was preplanned to look at this. I think stating the obvious when you're developing CD47, looking at CD47 expression is important. We did not specify a cutoff. Typically, when you think about targeted oncology, you need to do the all-comer work to understand what specific cutoff is the right one. We did not pre-specify 10%. That's why it was really important to look at a variety of different cutoffs to determine what was robust. It worked across multiple different cuts. For the breast study, we have not predefined the right cutoff. We want to get more data before we choose.

There are no additional questions at this time. 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. I appreciate the engagement, as always, and good questions. We're looking forward to future updates on this data as well as others over the next few months. Thanks again for the time this afternoon. I really appreciate it.

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