All right. Okay, thank you. All right. Well, good afternoon, everyone. Thank you for joining us on our next session. I'm Michael Yee, Global Head of Biotech Research here at UBS, and I'm very pleased to have members of the ALX Oncology team with us, including Jason Lettmann, Chief Executive Officer and Barbara Klencke, Chief Medical Officer. Couple of things I would love to start off. Jason, you just reported earnings, not because we about revenue and EPS, but because
Not yet. Soon.
talking about That's right. Talking about the pipeline and the developments and how you're going there. There are two programs we want to talk about today. Obviously, evorpacept, which is at CD47 for breast cancer. We talk about that because there's a major phase II readout that will be coming next year in a multiple indication. Also something that people are not as familiar with, which is ALX2004, which is an EGFR drug conjugate, which is data soon.
Yep.
Maybe I could just start with something new, which is this data coming up. I want you to maybe set the stage for EGFR antibody drug conjugates, something that people are not as familiar with. I would argue that antibody drug conjugates are some of the hottest technologies in oncology right now, so this could be really attractive, but people are still learning about EGFR.
Yeah, sure.
Data's coming very, very soon.
Very soon, in December. We just updated that as part of the earnings. First of all, just thank you for having us.
Yeah.
We've had a great day here. It's been a great-
Great conference, yep.
I think we're excited about both programs. Both are going to have really important inflection points. We were just talking, our hope this time next year is we're on the doorstep of two registrational studies if we continue to execute, and I think are well on our way. To your point on ALX2004, we have data coming next month. This has been a long time coming for this program, I think, as you followed. We designed, created, built this antibody drug conjugate in-house at ALX. We've been working on this program since 2021. We kept it under wraps for many years, and then started in the clinic about 12, 13 months ago or so.
Okay.
This data that's coming will be reporting from our phase I study. I think as we've discussed with you before, we really have been encouraged by what we've seen on enrollment. We've been encouraged by the way we've been able to escalate in terms of our dose. As I think has been covered, EGFR has been a tough target for an ADC.
Yes.
It's expressed in healthy cells, just like in the tumor. On-target toxicity have been a challenge. The way that we designed ALX2004 was really to address that. We also spent a lot of time optimizing potency.
Specifically, if you go back and look at some of the prior EGFR antibody drug conjugates, I mean, let's be frank, EGFR has been a target that's been around for a long time, monoclonal antibodies, and that's great for binding EGFR and having anti-drug activity by dragging ADC or activities there. By using it as a target to bind cancer cells and bring a payload to it could be very common for hence the excitement of ADCs. But if you're going to hit EGFR in the skin, we're now going to have a lot of rash and brain toxicity and have a rash.
Right.
You believe that you have found an epitope along the EGFR that is only expressed in cancer cells, not in the skin. Can you just describe that?
Yeah. We think it's, and you're right, it's been really challenging. We know big pharma has had specific ADCs against EGFR that have failed. We really went to school on both the history of EGFR antibody development as well as those failures. One of the decisions we made was choosing matuzumab. Matuzumab was developed by big pharma to uniquely bind EGFR.
Where did this come from?
EMD Serono.
Okay.
They took this through the clinic and put a lot of time and money into this program. What they found in the clinic was that, in fact, they were not having the same skin-related issues that the other-
Okay
EGFRs, like cetuximab, saw. We thought that'd be the perfect epitope to use as an ADC. We thought that could have more preferential binding to tumor. We also know it internalizes at the same rate as cetuximab. It's really the perfect vehicle to deliver a payload.
In prior studies, the monoclonal antibody does not have the skin rash that the others have.
Correct.
You feel pretty good about that. Now we are going to drag an ADC and a payload to it, and we need to prove that with your construct. I think two things. One is, in a multiple ascending dose cohort, we want to have good safety and tolerability, meaning less rash. I am going to let you mention how much rash is okay versus other antibodies. Then two is, we like efficacy. Not having any rash is good, but if we are not shrinking tumors, that is no good either. Can you talk about the data coming up and set the stage for how many cohorts, what is a good side effect profile, meaning what amount of rash is okay, and I am saying okay, looks great versus comparators?
Yeah.
What is good response?
Yeah, sure. I can talk about how we have framed the program overall. Then, Barbara, maybe you can talk to kind of how we went through dose escalation.
Okay.
The most important thing for an ADC is to keep the patients on drug.
Okay.
And that's what we ultimately want to show. We want to show minimal dose discontinuations, terminations, et cetera. Of course, rash is one of the top contributors. Patients hate rash. It's a chronic sort of building toxicity that certainly can be dose-limiting. We want all of this theory to translate in the clinic and show a very low rate of rash and a rate that isn't going to lead to dose discontinuation.
How much is a low rate of rash?
I'd say sub, I don't know, 15%-20%.
That's quite low. We know that cetuximab is in the 70% or above range.
Yeah, I was thinking more grade three and above and that would be
Okay, so, I do not have it right in front of me, but if I pull up the cetuximab label, total rash versus, I agree with you, grade three is also a very concerning rash. But what is the total rash and what is grade three rash for cetuximab ?
I mean, the total rash rate is quite high. I mean, that's
Like 80%.
70%, 80%.
70%, 80%, yeah.
Right.
Okay. We want to reduce that. Reduce the total rash is good too. What is a good reduction? Like cutting that in half?
I think that is a great target. Again, ultimately, we want to be able to show that we are able to keep patients on treatment.
Okay.
That's true.
Because we do want the payload to come in too.
Exactly.
Okay. In terms of grade three rash, I don't know the number off the top of my head, but what does cetuximab have in grade three?
I mean, I'd have to look that up too.
I think it's probably single digits.
Okay.
Yeah.
I mean, the grade three rash rate isn't terribly high.
Okay. Yeah, but overall, the rash causes people to drop out.
They're uncomfortable.
Grade three is definitely bad. If you are saying it is like 9%
They are unhappy.
in low single digits, that is remarkable. I am not saying there is zero, but you are saying And why would there be a much rash? Because there shouldn't be this epitope on wild type. Can you explain that?
I mean, I think that is what we saw. What is really encouraging and a good setup for what we are seeing in the clinic is that we did not see rash in the NHP work we did. We know from the early work that cetuximab and panitumumab did, that they saw that in the primate work.
Okay.
I think that's a good leading indicator for us that we effectively really reduced that challenge.
Non-human primates are not humans, so rash is a fairly reasonable model. Mice are definitely worse models for rash. The bottom line is, given that the data's coming in the next few weeks
significant reductions off the 78%, [inaudible] is good. The point about it is not because we're coming up with a drug that has less rash, it's to keep patients on drug so that we can see responses.
Correct.
Particularly with an antibody drug conjugate, not a monoclonal.
Yeah.
Because an antibody drug conjugate is going to be more potent on efficacy.
Yeah.
Keep them on drug and bring in the payload. So what is a good? How much response rates are we going to get?
I mean, I think it is a great point just on potency. Again, we designed, built this ADC in-house, and we spent a lot of time optimizing the linker payload as well as the epitope to drive both on-target delivery to tumor as well as the bystander effect.
Okay.
It is those two things that really drive efficacy in an ADC.
Okay.
When we shared earnings this week, I think we also just upped the expectations on numbers of patients.
Okay, go ahead with the numbers.
We now are guiding towards more like 35 + patients across this study, with 15 or more at the right doses in the right tumor types. In the past, we have talked about the study design being focused on four tumor types.
Yeah.
Head and neck, lung, esophageal, and CRC. I think internally we had identified non-small cell lung and head and neck as being the best opportunities. Once we got into the therapeutic range around four, we really focused on enrolling patients in those two tumor types. That is going to come through
That was 4s in four cohorts?
Four mgs per kg.
Four mgs per kg, right.
Right.
Which, philosophically speaking, is plenty of drug.
Yes.
You also would dose up because many times in cancer, you dose up to a dose-limiting toxicity. I mean, the general rule in cancer is give them as much as possible until they can't tolerate it because of maximal efficacy.
Right.
Would a dose-limiting toxicity be the reason you stop at four, or you stop at four because that's plenty of drug?
Yeah, I mean, I think with dose escalation, we're trying to take it up to a point where we'll see DLTs, right? I mean, ultimately
I am saying, yeah.
That is what we want to see.
What would be a DLT for this drug? Would it be the payload or would it be the?
It is much more likely to be payload.
Yeah
related.
Yeah.
Sorry, the strength.
Bottom line is, you are going to four mg per kg. The DLT would be related because of the payload. That is expected. Good news is, it is tolerable. People are not dropping off the drug, and it is plenty of drug to see responses.
Right.
Now, set the stage then for how much response rate should we get in those expansion cohorts, because you are at a high dose, we want to see responses.
Yeah, I think, here, we enrolled this study 100% in the U.S.
Okay.
This is top U.S. academic centers, so think MD Anderson Cancer Center , Moffitt Cancer Center, and the like. Then importantly, a very late-line population. This is on average a fourth-line patient population. When you think about the therapies available both in head and neck and lung, it is checkpoints, it is lines of chemo, it is prior EGFR. We took all of those patients.
When you say fourth line, you are saying that is like the average?
The median.
The median number of treatments in lung cancer and head and neck is four.
This will be, so three prior lines.
Three prior lines.
And through our study, they will be-
They are the fourth. I mean, I know the drugs in head and neck and lung cancer. They are not any great options in fourth line.
No.
Docetaxel, a chemo in second-line lung, barely gets you 15% response rate.
Right.
That's in second-line.
Right.
You're talking fourth-line.
Right.
I mean, there's nothing else out there. So, appreciated they're very heavily pre-treated, and I guess I would say that double-digit responses rates would be remarkable.
Yeah.
But maybe you could set the stage. I am also not saying that that means the drug does not work. I am just saying the higher the better.
I think that is right. I think from a benchmark perspective, double digits would be-
That is true.
remarkable.
Yeah.
I think greater than 30% would be outstanding, and anything above that would be transformational.
I mean, back me off here. I mean, at 30% response rates in fourth line, why isn't 30% response rates in fourth line? That is definitely remarkable.
That's what we're going for. Again, this ADC was designed for potency, and I think we're excited about
Serious, I never see 30% response rates in fourth line anything.
What's most exciting too is where we're ultimately going, of course, is not fourth line. I mean, the next step-
Sure
the step that we're actively taking is in the second line. Second line head and neck, and second line wild type lung.
Absolutely.
There's big opportunities for us in both.
Okay. We'll look for the response rates. We'll look for lowered rash. Patients need to stay on. If you have response rates, double digits great. I'd be shocked if I saw 30%, and that's great. Obviously, we go forward, which sounds promising. Now you want expansion cohort. You're going to open what? A big expansion cohort in lung and head and neck?
Yep.
That would be cool.
Second-line lung. I think we'll look at both EGFR experiences amongst EGFR naive.
Yeah.
That standard of care is changing. As we know, petosemtamab is likely to get approved in frontline, which is an EGFR bispecific, so we're going to want to understand our activity both in patients that have not seen an EGFR and those that have. We'll have a somewhat of a signal.
I'm sorry. Frame that again. Obviously, second-line EGFR, your point is that they're I mean, I know the first-line agents. You're saying that something else is coming in front.
Well, Peto, which is the Merus molecule, and then Genmab.
Right, for head and neck.
Head and neck, they had frontline-
Yes
data that suggests that they will be moving frontline.
Yes, they just reported that data.
What is really important and it is not defined is what happens post-petosemtamab.
Right.
A lot of the data sets to date have been largely EGFR naive patients.
Right.
It is going to be important for us to do both, and one of the things that we think is particularly exciting about our choice of epitope is that it is different, because petosemtamab and all the other bispecifics use cetuximab
whereas we use matuzumab.
We know that some of these patients who progress will mutate off of that binding site.
Is the payload the same? I guess it is not a concern. This is about bispecifics.
Yep.
You are not worried about that.
Right.
It is a different situation.
That is what is interesting for us in head and neck. We have the potential not only to bring the unique epitope advantage, but to be the first topo in that field.
I think that is why we need to move quickly.
Okay. So in head and neck, you would want to run EGFR naive and EGFR experienced, and see if there is any difference. The point is it could work very well on both because you have a different epitope. So as long as the antibody drags through the epitope, then it should bind and therefore drag the payload in.
Correct.
That is all that is needed.
Yep.
As long as there's not like a loss of the-
Right
of the epitope.
Right.
Okay. In lung cancer, what would be a plan there if the data are good?
I mean, wild type lung is where there's the most opportunity. In mutant lung, there's other topo ADCs that have advanced. But in wild type lung, particularly in the squamous setting, there's really nothing available. DATROWAY as well as the sac-TMT data has been relatively disappointing in squamous lung. What's so interesting for us with an EGFR is we know in squamous that EGFR is very highly expressed, and that's a big unmet need. In this data set that we'll be reporting, we're going to have a number of squamous and non-squamous patients. But that wild type setting is where we're focused the most.
I need to fact check myself because I feel that sac-TMT has had very positive data in combination with chemo in PD-L1 negative patients.
Across PD-L1 positive, obviously with PD-L1. But sac-TMT definitely looks good. I covered Merck. I think it is good for first line. There is certainly minimal options in second line. Once again, sac-TMT is probably going to be dose attenuated in second line. They do not care about second line. They want to go first. But there is definitely room in second line. My question is if it is using the same payload, there is a small chance of mutations that would cause a problem. But the bar is open there, what you are seeing in second line.
Yeah, I mean, most of those studies you mentioned are on the mutant side of things.
Yeah.
And, you know, we certainly understand the challenge of bringing another topo. So, you know, a topo after topo, I think the jury is still out, at least with our drug. We do have a different epitope, so that could be-
Okay
a way that we could play there. But for us, on the wild type side, there's much less data.
Okay.
There's much more open field, including versus sac-TMT.
Okay.
That's where we see a big-
Yeah. The reason is because obviously they were doing that to hit mutant because they didn't want the on-target toxicity, or what was the-
I think the biology is slightly better in the mutant setting.
Yeah.
I think the DATROWAY data was also pretty surprising to folks. I think they expected it to work as well at squamous as it did in the non-squamous patients, and they didn't see that.
Okay.
They saw, a relatively low ORR.
Your point is in squams that the other drugs have not done so well.
They have not.
That is certainly true for Dato-DXd.
Yep.
It did not ultimately get approved, I think, in that setting.
Right.
Now we're waiting on AVANZAR and some other stuff. But sac-TMT, I feel maybe better, but this is a different question. Plenty of open room.
Right
for that. Okay.
Yeah.
Anything else to say about that? Otherwise, we look forward to this data in December.
Yep.
I guess, as every company here has said, who still has data remaining this year, they say not during Christmas week.
Not during that.
Some people, like Biogen, say first half of December. I mean, Arcus Biosciences was just here. They said it's coming very soon.
I don't know when our holiday party is. I think it's the 21st or something.
Okay.
It's definitely going to be before that.
Definitely before your holiday party. Okay. Well, I'm hoping for a good holiday party.
We are too.
Okay. Let's talk about, in the last 10 minutes, something that is very important and there's a lot of data. Here, let's walk through this because I think it takes some explanation, and once people understand it looks pretty interesting. So, evorpacept, it's in metastatic breast cancer, HER2-positive metastatic breast cancer. Can you talk about the ongoing phase II study? Explain that, why it's designed like that, and then we'll talk to your update, which is something that just happened.
Sure. Barb can walk through the study design. I think with EVO, what is so compelling about where the program sits right now is we have the opportunity to be the first, and frankly only, option for patients that are CD47 high. That's what's really changed in our story, is an ability to go from being this broad sort of IO approach to a very targeted.
CD47 and perhaps, because in combination, any anti-CD16 antibody that would have.
Any HER2 antibody. That's exactly how-
My point was that if it was a different, like in colorectal cancer, as long as you had a CD16 antibody, this would be with it.
Yeah.
This is like proof of concept.
Any FC- active antibody-
Yes, exactly.
our drug will work in combination with. In this case, we use Herceptin.
Yep.
That's what we've learned, is that when we see CD47 and we see HER2 expression, our drug is working.
Okay
phenomenally well. Do you want to touch on the trial design?
Yeah. ASPEN-09 is the ongoing trial. We had great data in the past. We did a relatively large randomized phase II in gastric on the basis of a trastuzumab-based regimen, added evorpacept, and we saw a good increase in response and durability of outcome. We have a smaller data set in breast cancer with zanidatamab. Now what we want to do is enlarge our experience in breast cancer prior to a phase III. We are enrolling 120 patients. They will be HER2 positive based on historic archival. Then we want to look very closely at CD47 to understand where the best threshold is. The cut point that defines CD47 overexpression needs to be optimized before we go into a phase III.
We'll look at our outcome data next year, second half of 2027, and what we're hoping to see is a high response rate in patients who are CD47 overexpressing, the response rate in patients who are HER2 negative on fresh biopsy, or CD47 negative. So kind of an internal control within the study.
Yeah. Let's walk through that. You plan to enroll 120 HER2 positive second line plus breast cancer patients. If you get 120, the idea is that you will check to see if they have a recent positive HER2 positivity. Is that a fair way to So it's 120 patients because they're going to get in the study, and the files show that they were HER2 positive at one point. Is that because there's a second biopsy taken, or how do you confirm fresh HER2 positivity?
Yeah, both. We really emphasize the importance of sites, investigators getting fresh tissue. But no matter what, we'll take the tissue that they have and we'll look centrally. We're able to confirm HER2 positivity in about We haven't talked about our data yet, but in general, what you see is somewhere in the range of 60%-70% of HER2 positive patients can be confirmed-
Right
on central read.
Let me clarify. Not because they are getting a fresh Are they getting a fresh biopsy?
We have encouraged our patients to do so, but a portion of our study population will have had a recent biopsy, and another portion will have had their archival tissue.
And if you use the archival, how do you confirm that they are still HER2 positive?
We also are taking ctDNA. The ctDNA is blood-based, of course, and that has a relatively high-
Okay
concordance rate in terms of-
Okay
positivity.
When you enroll 120 patients, and I am going to get to the two cut points, the population first has to be HER2 positivity based on fresh biopsy or based on ctDNA if they did not have a fresh biopsy, and that is basically 90% concordance.
Okay. Now, when we get that, you think how many patients of the 120 will be HER2 positive?
I think it's probably around half.
Okay.
Maybe a little bit higher.
Okay.
Again, it's in flux because if you think about when we started enrollment to now, we have a lot of-
Now, I see you said 60%-70% historically have HER2 positivity. Then when you take people with fresh biopsy, they are definitely going to be HER2 positive. Otherwise, they would not have gotten into the study. That is fair.
Well, some of the patients. Well, we are-
Are we enrolling patients, then taking the fresh biopsy, and they are not HER2, but we let them stay in the study?
The local read would always have to be HER2 positive. There is also the discordance of local versus central confirmation on the same biopsy sample is not always confirmed.
But our goal is to make sure that the patients that are in the HER2-positive bucket, if you will-
Yeah
are really HER2-positive.
Okay. Yes.
Because ultimately, that's how our drug hits.
Okay. Without getting into the details of that, once you test either for those, it is about 50% of those are likely to be HER2-positive.
Now, when we get to those 60 out of 120, this is the critical part. You are seeking to test this drug particularly in CD47 high patients, because we want HER2 positivity. Otherwise, obviously, the backbone, Herceptin, is not going to be doing anything to obviously bring ADCC in with a CD47 antibody to work. Makes sense. So, what percent of the patients are going to have CD47 high versus low?
It is probably around 50%-70%, depending on how we define the cutoff.
Now, when I recall the gastric data, I recall that there wasn't a huge dramatic difference in response rates. I thought it was anything above 5% or something. Help me out with the numbers. Basically, have pretty good response rates. So, I wouldn't want to cut the cut point too low, because we don't want to cut off patients in the market opportunity.
Yeah, exactly. What we showed in the gastric study, which was a randomized study, is there wasn't a lot of magic in where we set the cut point.
Right.
The signal was so strong.
that you could have just said they have CD47 versus don't, and that in and of itself would drive a really strong drug effect.
Okay.
If that happens here, now it's still important to define the cutoff, because that's going to help inform the commercial opportunity.
Okay.
But ultimately, when you see that sort of treatment effect, these things get-
This is a phase II, so let's not get too much into the worms of, "Oh, it's post-hoc analysis. You looked at those above 5%, but those above 10% really had it, but those above 5%." I don't want to get into that. Just to be clear, the primary endpoint is response rate in HER2-confirmed positive for the patients and CD47 high. But I ask that question, what is the protocol as defined as CD47 high?
We've defined it as sort of on/off, if you will. So, CD47 presence versus not.
And presence is what, above 5%?
Yeah.
Based on the gastric data, above 5%, there was a response.
Right. Based off of any data as well.
Okay. What percent? It is about 50%-70%.
Right.
If I get it right, with CD47 above 5%.
Yeah.
So, it takes 120, cut it to 60, because those are people who actually are clear to have HER2 positivity, then 50% of 60 is 30 to get those who are CD47 high. We should have 30 people who are CD47 high, 30 that are CD47 low.
Right.
In CD47 high, what type of response rate would you expect? Then in the low, what is the background? See, it is very interesting, because you almost kind of serve as your own control. Think about that, because in the CD47 low, they basically don't have any CD47, so your drug, which obviously blocks this "they don't eat me," is not really going to draw any difference, that was sort of seen in the gastric data. It sort of serves as its own self-control, because retreating with Herceptin ain't going to do much if you've already failed in HER2. So, that should be like, what, 10% response rates, not even that.
That is exactly why we designed the study like this.
Self-control.
Yeah. If you look at the standard of care post in HER2 in the front line, it is yet to be defined. We do not know. I think when we look at the real-world data, it is abysmal. There was a real-world study out of Asia that was published at the last ESMO, about 600 patients, and they showed 14%-15% ORR with three or four months PFS. We know that the bar, so to speak, is low.
Yeah. It is 15%, again, with a retreatment of a HER2 drug, it is a retreatment, on someone who has already failed basically in HER2.
You give them Herceptin, they might have 15% response rates. You are definitely shooting for a lot higher than 15% if you throw in a war machine.
The other thing interesting about that real-world study is it did not matter what they got. They could have received a TKI or not a Herceptin, it was still very low ORR. If we can double that, I think we add something. That is how we are thinking about our-
That would be 30 patients. If you have 30% response rates, it is basically double. That is definitely above anything anyone has seen before.
Right.
Double the control arm of 15% around self-control, then that would be pretty obvious. It is open label, but again, this is response rates. There is no placebo effect. That is pretty real.
Yeah.
Okay.
Yep.
Fantastic. All right. This data will come in the second half of 2027. Will you give updates on enrollment? Because just to be clear, you just gave an update on last week's earnings call, and you said, "We want to go to the full 120. That is what we want to get to. That is the most robust readout.
Yeah. I think we feel really good about our enrollment projection. I think that guidance is right on.
Where are you with cash, Jason?
We have cash through [inaudible].
[inaudible].
Raised $159 million and continue to be well-funded to get through these key milestones.
to our milestones that we are. Yep.
Fantastic. Thank you very much, guys.
Thank you so much.
Appreciate it.
Really appreciate it.[inaudible]
Jason, that was great. Thank you for-