Good afternoon, and thanks for joining us to have a conversation with Jason Lettmann, Chief Executive Officer, and Harish Shantharam, CFO of ALX Oncology. ALX Oncology is a clinical-stage immuno-oncology company advancing two internally developed programs towards multiple inflection points in 2026 and 2027. The lead program, evorpacept, is a first-class CD47 blocker designed to separate blocking the "don't eat me" signal from delivering the "eat me" signal through a partner Fc active antibody. CD47 expression has now emerged as a predictive biomarker across both ASPEN-06, the gastric study, and also the HER2-positive breast cancer study. That thesis is also being tested out in ASPEN-09, a phase II study with evorpacept plus trastuzumab and physician's choice of chemotherapy in HER2 experienced, HER2-positive breast cancer, with top line data expected in mid-2027.
The second program, which is the more interesting program, an exciting one, the novel EGFR-directed TOP1i ADC, ALX2004, is currently in phase I dose escalation across four high EGFR expressing tumor types with safety data and dose escalation data expected in the second half of this year. To discuss both the CD47 as well as the EGFR ADC, let's get started with Jason and Harish.
Great. Thank you. Thanks for having us.
For those who are new to the company, CD47 carries a lot of history. There was a $5 billion acquisition at one point. There was a $2 billion acquisition that really didn't work. Why do you think evorpacept mechanistically is different and is a better bet on the same target?
Yeah, sure. I think CD47, like a lot of things in biotech, there's been all sorts of ups and downs. I think you could write a novel about the history of drug development in that space. That's what makes it fun. What we get to do each day is it's unpredictable. As we like to say, it's not the target's fault, however, that things played out like they did. I think there's just been no question over the last 15 or 20 years of research that one of the fundamental ways in which cancer evades the immune system is via CD47. It's a tough target because it's also how our healthy cells signal to our immune cells to be left alone.
Because of that challenge, that fundamental challenge of expression, where you have expression on healthy and you have expression on cancer, again, that's what required a unique approach. If you think back to our company's founding, there was really two schools of thought. One, the conventional one, was let's design an Fc active antibody, one antibody that will both activate, drive "eat me," and then block "don't eat me." What that ended up doing was causing all sorts of significant on-target issues because of the fact that CD47 is so widely expressed, they had really significant on-target toxicities that really weren't apparent until after those acquisitions when they ran the randomized studies.
Okay.
That for us, was what we tried to solve at the company's founding. We bifurcate the "eat me" signal. We have a dead Fc, so we provide full blockade of the "don't eat me" signal, and then we use an Fc- active antibody like Herceptin, for example, to provide the activation. That mechanism is the same that we're pursuing today, and it's been a beautiful thing to watch that translate in the clinic. We are far, far safer after dosing over 800 patients here. We know it's safer, and frankly, quite safe in general, and we know that the activity has been quite strong when we get that combination right as well.
Okay. One of the goals internally for you is to get both of these molecules into registration level studies in 12 months from now. For that to happen in each of these programs, what needs to get done between now and then?
Yeah, we need to be working hard. As much fun as investor meetings are, Harish and I get to do this, but everyone else back at the office is doing the real work. I didn't mention our second program that you alluded to, which is ALX2004, which is our EGFR-targeted ADC. We're super proud of that molecule. We designed it in-house by a really great group of world-class chemists and protein engineers. Since 2021, we've been pursuing that in-house. We started dosing patients about a year ago. That's been a very competitive space and also a very challenging one. Targeting EGFR with an ADC has been tough. We think we've cracked the code, as we'll get into. We have a novel epitope called matuzumab that we're using. We think we're the only ADC in development that utilizes that backbone.
And we think our linker payload architecture is quite unique, too. As you mentioned, our goal internally is by the time we're sitting here next year is to have both programs on the doorstep of a phase III. And I think so far so good. Execution has been really quite strong across both.
Let's dig a little bit into that 2004 program. Currently, you're in the dose escalation phase, and you're going to see that data later this year.
What does that update need to show for you to declare that you got the right therapeutic window and can move this molecule forward?
Yeah. I think the program's progressing quite well. We started dosing at 1 mg per kg. We then doubled to 2 mg per kg, and we've doubled to 4 mg per kg, and now we see signals over the summer that we're now beyond 4 mg per kg , and I think that's really important. For most TOP1i-based ADCs, 4 mg per kg and above has been in and around the therapeutic window. I think first things first, we want to show that we have a therapeutic window to play with, so to speak. I think what's really key to remember about this study is that it is a U.S.-only site.
Our sites are only in the U.S., so we are enrolling and treating patients at the top academic centers, so think Moffitt, MD Anderson, et cetera. Because of that, we're seeing late-line patients, so on average, our patients have seen three or four therapies before entering our study. It's a great place in which to test your drug, and it's a high bar. I think with this update, we really do want to define and I think lay out where we see the therapeutic window, but also hopefully show some signs of activity as well.
Okay. The other uniqueness about this ADC is it's an epitope that has not been tested by anybody else, and it was originally developed by Merck Serono to avoid EGFR-related toxicities.
Right
Beyond tolerability, is there anything unique about this epitope, and do you think this epitope could be used in other EGFR-driven diseases?
Yeah. I think through the development of ADCs, it's been pretty clear that the choice of epitope is really important, and our team went to school on the history of EGFR development to come to this epitope. Cetuximab and panitumumab are both great drugs. The but is their on-target toxicity, so significant skin blistering, derm-related toxicities that have been tough. I think our insight was if you strap a payload to that antibody, you could really open up yourself to more on-target toxicities. Matuzumab, as you mentioned, was developed by Merck Serono-
Yeah
...with a unique binding pocket, with a unique domain, that after a whole lot of screening, they determined did not have those issues, and we think is selective to tumor in that way. We did not see on-target related skin toxicities in our NHP work, and that's what we're trying to show in the clinic. You mentioned potency. I think the other really clever approach our team took was to take our linker payload, synthesize 60 different linker payloads, and then throw them up against ENHERTU as our benchmark. All of our benchmarking work preclinical was versus ENHERTU, and that's why we think we've also developed an incredibly potent molecule.
Okay. The other toxicity which is normally thought of is the interstitial lung disease, which you don't see with your molecule. Also, in terms of the publicly known benchmark is a CSPC's molecule, that SYS6010
which has been escalated up to 6.4 mg per kg , but they do see some all grade rash in there, but no ILD there. Is that the real molecule, and is that the real data that you should be held to when the data comes out, or that's not the right
Well, I should ask you what you are going to hold us to. But I think for us, I think it is a good benchmark. As I mentioned before, rash, ILD, has in the past translated. If you see that in your NHP work, there is a very good chance you are going to see it in the clinic. We did not see that in our NHP studies, which I think is important and a leading indicator, we hope, to the clinic. There are certainly other EGFRs in development. The CSPC molecule is one of them. I think what they have demonstrated is the same as what other TOP1is have seen, which is very challenging to push dose beyond 6 mgs per kg.
Although they tested that, they ended up settling on a dose between 4.2 mg per kg and 4.8 mg per kg, and that is typically the right window, so between 4 mg per kg and 5 mg per kg is where a lot of these other ADCs have settled out in terms of dose.
We know that you are looking at four different tumors with this initial study, and let us say we hit the jackpot and get good data on all of the four indications.
Is that your expectation? Because I might not be doing a good job of setting expectations.
Are there any favorites of the four kids?
I have three kids, and I have learned to not pick favorites, although all parents do have one, I would say. No, I think what is exciting about all four, we are going after lung, head and neck, esophageal, squamous, and colorectal. I think of those, of course, colorectal has been very hard, and it is very hard in dose escalation because you know the bar is very low, but it is challenging, right? If you see two out of 10 patients in your colorectal cohort, is that good? It is tough. I do not know. I think for us, head and neck, lung, and esophageal are where we are most focused on at this point, and I think we have tried to think about those things in dose escalation.
We have been pretty conscious about trying to tilt more towards head and neck and lung as we enroll these patients, and we continue to feel like the opportunity there is significant for our ADC.
Okay. Moving on to evorpacept and especially on the ASPEN-09 study. This is a single all comers, and you are looking at 80 to 120 patients at this point.
The expectation is to see some data in mid-2027. Can you just describe for us a little bit about the study itself and how sure are you that this trial could give us some data or the first set of data in mid-2027?
Yeah, sure. The setup for this study is a lot of clinical data. Really what informs the goal here is our two prior HER2-positive studies. The first, which was a global randomized study in gastric, where we showed an almost 40% delta in terms of ORR versus control in the CD47 high population, and a tail that looks like an IO tail, right?
We had a durability of response of about 25 months and a PFS hazard ratio of 0.39. Just an incredibly strong benefit versus control. Then we went on and did a study with our partners at Jazz Pharmaceuticals, combining evorpacept with zanidatamab.
Also in HER2-positive tumor, this time in breast. Showed a really compelling benefit there. From there, felt like breast was the right place for us to go. Now we're enrolling ASPEN-09-Breast, which is focused on, this will be the third study in HER2-positive tumor type, looking at patients that have progressed on an HER2, so second-line or greater patients. Here, I think we're hoping to show similar. Again, we went five for five in our first-
Yep
...study. I do not think we are going to show 100%, so hopefully that will not be your expectation. But I think we saw a really compelling benefit. One of the things that has really changed in our story is just the fact that CD47 is very clearly a predictive and strong biomarker for response for our drug. I think in IO, one of the biggest challenges and why there has been so many failures has been the lack of a biomarker.
The PD-1 story is so strong and has benefited so many patients because we know if you have a CPS score of call it greater than one, you are going to do great on pembrolizumab. A lot of these other attempts in IO have failed due to the lack of that, in my opinion. I think the fact that we now know that patients that overexpress CD47 really have a significant response to our drug has been a game-changing, I think, insight for our program. That is what we are focused on. Here we are going to take all comers, so we are going to have patients that are CD47 high and low. Although it is a single arm study, we are going to have a built-in way to understand which patients do best.
With that, the goal is to inform our phase III, which we hope to be pushing forward late next year.
Okay. Is there a threshold, an internal threshold that you folks have when you are looking at the study to make sure that you are embarking on the right trajectory there?
Yeah. The unfortunate thing in this patient population in metastatic breast is for patients that progress on an HER2, the prognosis is really quite poor.
Yeah.
At last year's ESMO, there was a really nice real-world study of about 600 women, and it looked at progression, and it is really abysmal.
Yeah.
It was about a 15% ORR, 3-4 month median PFS. That is the bar, that is the need, that is why we are doing what we are doing. Again, I think if we can show certainly 30%, which would be a doubling of ORR, that would be fantastic. On our last KOL call, I think Dr. Sara Hurvitz thought 6 months-ish PFS would be the number to hit. Again, not an exceedingly high bar, but there are a lot of patients out there in need, and it is a really significant population that we are trying to address.
Okay. Let us see if we can answer this question.
Okay.
Because you haven't given the threshold on this, which is the CD47 expression cutoff. What's the expression cutoff that you would like to see? How will that help you plan the next study?
Yeah. The reason we're not answering that is that we don't know. I think we do know a fair amount about how we think it should look. If you look at our gastric data, one of the most pointed questions that I got from investors when we first shared it is, "Wow, this looks really quite cute. Nice work on your biomarker-defined population, but did you just cherry-pick one cutoff that looked good?" We shared later at ESMO the fact that you could have just about picked any cutoff in terms of IHC score, and the response still held. When we got the data from Jazz Pharmaceuticals with our study with zanidatamab, we were really quite encouraged that if you did nothing more than just look at patients that had no expression of CD47 versus patients that did, that was enough to drive the five out of five.
Again, I think when you're looking at a biomarker-defined therapeutic, it's all about the spread, the magnitude of benefit, and given what we've seen in the past, if that carries through, we're going to have a lot of options when it comes to defining the specific cutoff.
Also, you're working with Roche to set up the diagnostics for this.
In terms of what sort of data do you need to see to make sure that you get to set the diagnostic in place, and do you have to have the diagnostic in place before you start the phase III study?
We do. I think we have to have the work done, and so we've been investing heavily with Roche to make that a reality, to push forward a companion diagnostic. I think the good news here is that there's no magic into the test itself, right? We're looking at a target that's expressed on the surface of a tumor cell. We're using standard IHC techniques to do that. So again, there's no magic or discovery that needs to happen. We just need to put in place the rigor of that test and then deliver more data to help support what cutoff we'll use.
One more question on the ASPEN-09. I know you plan to talk with the FDA before you get going on the phase III study. There's also some thought about having to do with a smaller patient population because of the combination that you're going to test. Can you describe what is it that makes you feel that you can do a smaller patient population than running a 600 or 700 patient study?
Yeah, I think it's really one of the positive attributes of running a biomarker-driven study. You usually can end up targeting a smaller patient population. Again, it goes back to the magnitude of benefit.
Like I mentioned in the gastric study, we showed a hazard ratio of 0.4 better across all those different cutoffs. Of course you never power a phase III at a hazard of 0.4, but it does give you, I think, the ability to run a much tighter phase III. We'll see. Ultimately, FDA will be informative and be a big driver of that decision. But we do think if you look at the benchmarks in the history, you have an ability to run a much tighter phase III path.
So, you also talk about how evorpacept can be a natural combination partner for any Fc- active construct. When you say that, what are you thinking about in terms of a combination and to develop a combination product, do you need to have a partner, or you can get started on the studies?
Yeah, I think in the dare to dream scenario of what if this works, right? The what if this works is we're sitting on the next big immune checkpoint, and there's not a tumor type, including in a heme malignancy that doesn't overexpress CD47.
Because of all the funkiness of how this played out, it just so happens we're the last one standing. When we think about the competitive setup of potentially owning, and owning should be in quotes if we had a lawyer here, but owning a space where we are the only CD47 game in town for patients that overexpress CD47. I think our combination potential is really endless. Any bispecific or Fc- active antibody we could play with. I think that's where, yeah, that's the kind of dream and vision of what we're pursuing.
Okay. Harish, you closed the quarter with $153 million, and what sort of a runway could you get from that and some of the things that Jason is talking about and how much of that can be accomplished with what you have?
Our cash runway guidance gets us in through the first half of 2028. We've been singularly focused on executing the two trials that Jason alluded to, the phase II trial of evorpacept through the next major catalyst next year, and then the ALX2004 ongoing phase I escalation study. We think there's going to provide some critical data sets that'll help us propel towards what you had initially started alluded to, is getting it ready for a pivotal study ready. We're also ensuring that we'll be ready when the data supports it with the investment in CDx and some CMC, along the way as well. We feel good with the balance sheet we have and our ability to execute the ongoing trials.
Just to close out, Jason, beyond these two molecules, is there anything in the pipeline that you would unveil in the next 12 to 18 months?
Yeah. I think, again, we're really proud of the two programs, in-house developed programs, and we continue to have active research efforts. I think as was the case with 2004, from 2021 up until last summer, we didn't disclose that program publicly, and I think we'll continue that trend. Once we get closer to IND and have things that are more advanced, we look forward to sharing it at that time.
Great. Thank you very much, Jason.
Thank you. Appreciate it.
Thanks, Harish.
Thank you.
Okay.
Thank you.