ALX Oncology Holdings Inc. (ALXO) Earnings Call Transcript & Summary

Hey, good morning, everyone. My name is Daniel Wolle. I'm an analyst covering smid-cap biotech. Welcome to the 40th Annual JPM Healthcare Conference. I'm joined today with President and CEO, Jaume Pons, of ALX Oncology. As a reminder, if you want to ask questions, you can submit using the blue Ask a Question button to the question portal and I'll be able to ask those questions. Without further ado, Jaume?

Thank you, Daniel. After the customary disclaimers on Slide #1, let's move to Slide #2. Our main focus is evorpacept, a highly differentiated CD47 blocker that is now in Phase II clinical trials for 3 solid tumor indications and ASPEN early clinical studies in 2 heme indications. Evorpacept is designed to be used in combination with a long list of anticancer therapies. So if we are successful, it can be a cornerstone of therapy. Evorpacept was designed to be used on top of the standard of care in combination. Continuing with the CD47 pathway, we have a CPG conjugated SIRPalpha antibody in IND-enabling studies in collaboration with Tallac Therapeutics. This is a CD47 blocker that directly activates dendritic cells using CPG as an agonistic molecule. This molecule will jump start the immune system against cancer. This molecule is designed to have a very strong single-agent activity to be used in combination and as a single agent when a standard of care fails. In addition, we have early pipeline of cancer-activated antibody that conjugates that works as stand-alone development and also, we'll be able to combine with evorpacept. Thanks to investors, also we are in a very good cash position with $385 million as of September 30, 2021, with enough cash into 2024. And that includes all the randomized studies that we are now trialing with evorpacept. On Slide #3, focusing on evorpacept. We have promising data in combination with antibodies, checkpoint moderators and chemotherapy. It is the only CD47 blocker that has shown positive data both in solid tumors and hematological malignancies. Evorpacept has a very good safety profile that enables high dosing and flexible dosing to match a combination with the drug that we are going to combine with. The clinical trials that we are now -- running now are selected by mechanism. For example, in gastric cancer, we're combining with Herceptin to test the mechanism of combination with anticancer antibodies. In head and neck, we're combining with KEYTRUDA to test the mechanism of combination with checkpoint moderators. In MDS, we're combining with azacitidine and also in NHL with rituximab. From here, if these studies are successful, we can branch out into many other studies in other combinations and other indications using the same mechanism. And from there, also we can go to other modalities like antibody drug conjugates, et cetera. On Slide #4, you have a snapshot of our clinical pipeline. We ran a very large Phase I study, testing the different mechanisms. And now we're running Phase IIs in mechanisms that we think evorpacept has shown activity. For example, based on our promising data in Phase I and head and neck in combination with KEYTRUDA, we are now running 2 Phase II studies, the ASPEN-03 and ASPEN-04, ASPEN-03 in first line head and neck in combination with KEYTRUDA, ASPEN-04 in combination with KEYTRUDA, 5FU and platinum also in first line. We have Fast Track designation, and we are running these studies in collaboration with Merck. Also based on our positive data in gastric cancer in our Phase I study, we are now initiating one Phase II study. It's a randomized study that compares Herceptin, CYRAMZA, paclitaxel with evorpacept versus Herceptin, CYRAMZA, paclitaxel. We also have Fast Track designation for this indication, and we are running the study in collaboration with Lilly. Based on this positive data with Herceptin, we are now doing a collaboration with Zymeworks to test evorpacept with zanidatamab, a bispecific anti-HER2 antibody and this is on breast cancer high HER2, breast cancer low HER2 and low HER2 many other indications. In the hematological setting, our Phase I show very good activity in NHL in combination with rituximab, and now this has been developed as an IST in combination with rituximab and lenalidomide. Also we are running studies in MDS and AML and MDS in combination with azacitidine and AML in combination with azacitidine plus venetoclax. And based on this data, we're going to be moving into Phase II studies also randomized for these indications. Focusing on evorpacept, ALX148, and its mechanism on Slide #6, you have the CD47 is a target that is highly expressing cancer. Here in the left panel, you see the expression of CD47 in tumor in red, but also very highly expressed in normal tissues in black. Therefore, if we went to design a molecule that kills all CD47-positive cells, it's not going to have a very good therapeutic window. What's interesting to me about CD47 is its function as a checkpoint modulator, in this case, for myeloid cells. On the right panel, you can see that CD47 is highly expressed on cancer cells and interacts with SIRPalpha on macrophages. This is an inhibitory signal called a do not eat me signal. But macrophage to be fully activated requires a positive signal. This positive signal can come from the Fc interaction of antibody with Fc gamma receptors or other kinds of damage to cancer cells. What's important to remember here is that blockade of CD47 SIRPalpha by itself will not activate macrophages. That's very similar to PD-1, PD-L1 on T cells, where blockade of PD-1, PD-L1 is not efficacious unless the T cell already has T cell receptor that recognizes MHC peptide on the cancer cell. So this mechanism is equivalent to PD-1, PD-L1. But instead of T cells, we are now talking about myeloid cells, macrophages and dendritic cells. On Slide #7, this is an approach that we talk about modulating CD47 as a target. We wanted to design a molecule that highly potently blocks CD47 and prevents interaction with SIRPalpha but does not provide the required positive signal of the Fc gamma receptor interacting with Fc receptors. And why is that, it's because we do not want to target normal cells for destruction, for example, red blood cells. Evorpacept does find red blood cells, but because the Fc does not interact with Fc gamma receptors, these red blood cells are not been destroyed. And therefore, we don't have anemia, thrombocytopenia or neutropenia problems that other CD47s have shown in the clinic. And then because of this property, we can dose very high, fully block the CD47 pathway, and then in combination with an anticancer antibody or other anticancer therapies that specifically target cancer cells, we can target the destruction of the cancer cells specifically and sparing normal cells. On Slide #8, you have the design of evorpacept. We've painted it like an antibody, but it is not an antibody. It's a fusion protein that comprises a binding domain that is a high affinity SIRPalpha. So it's the extracellular domain of the receptor CD47 that has been affinity mature to bind CD47 with very high affinity, with picomolar affinity. The normal type SIRPalpha binds CD47 with micromolar affinity. So we have thousands of higher affinity than the natural SIRPalpha. That is fused to a completely inactivated Fc domain that does not interact at all with Fc gamma receptors. That maintains the interaction with Fc. So it has the same PK as an antibody, in our case, about 30 days half-life at the steady state. The molecular weight of this molecule is half of our antibody. Therefore, we have twice the number of molecules per gram. Our 10 mg per kg is equivalent to 20 mg per kg of an antibody. And that, of course, has a very significant astigmatic advantages. In addition, this smaller size has -- this smaller size may provide a better tumor penetration. What's quite unique about evorpacept is that we have cross reactivity to human, monkey and mouse. So our animal models are more relevant with the clinical setting than other CD47 molecules. The manufacturing of evorpacept is completely antibody standard with yields of 4.5 grams per liter, and now we have a stability of more than 2 years at 4 degrees. So it's a very well-behaved molecule. On Slide #9, you have the safety profile of evorpacept so far in combination with different agents. And you can see that across multiple studies, including chemotherapy, checkpoint modulators and anticancer antibodies, the safety profile has been really good. The most common side effect is fatigue, low-grade, some rash, also low-grade, and the other side effects are quite sporadic and single-digit and mostly low-grade. We have never reached maximum tolerated dose with evorpacept and we have dosed up to 60 mg per kg, which remember, that's equivalent to 120 mg per kg of antibody. Now let's move to different indications that evorpacept has been tested. In Slide #10, we are going to focus in evorpacept in HER2-positive gastric cancer. On Slide 11, you have the mechanism of evorpacept in this setting. In this case, cancer cells express HER2 that can be recognized by Herceptin. And this interaction with Fc gamma receptors would activate macrophages. However, the CD47 SIRPalpha interaction prevents activity of Herceptin in macrophages, by blocking CD47 with evorpacept, completely enable this mechanism of Herceptin that is usually not in use. On Slide #12 is our data of evorpacept in combination with Herceptin, CYRAMZA and paclitaxel. You can see in this study that we have a very high response rate of 72%, so most patients do respond, including patients with lower HER2-expression, for example, 2 plus. And you can see that the duration of response is also very long. Patients -- most patients respond, and many patients stay for a very long time in the study. You can see the overall survival of 12 months is 79%. These 2 numbers, response rate and OS of 12 months, are much better than the benchmark of CYRAMZA and paclitaxel of 28%. With this positive data now, we are moving into a Phase II study where we are comparing Herceptin, CYRAMZA, paclitaxel versus Herceptin, CYRAMZA and paclitaxel. These 2 completely demonstrate the activity of evorpacept in this setting. Remember that the standard of care is not Herceptin, CYRAMZA and paclitaxel. It is actually CYRAMZA and paclitaxel because it has been shown that Herceptin does not add anything in the second-line setting. If this study is positive, we're going to move into a Phase III where we're going to do Herceptin, CYRAMZA and paclitaxel versus the regulatory benchmark CYRAMZA and paclitaxel. The endpoints will be response rate, OS, PFS and duration of response. Next, I want to talk about evorpacept in head and neck, in this case, first line. On Slide #15, you have the mechanism of evorpacept in combination with checkpoint modulators. This mechanism is slightly different than done in the signal macrophages. In this case, the inhibitory function of CD47 is SIRPalpha expressed on dendritic cells. Dendritic cells also express SIRPalpha. And we have shown that they are completely inhibited constantly by the CD47 interaction. By blocking CD47, we activate dendritic cells and then we are able to present antigens to T-cells and then T-cells, when they are unleashed by blocking PD-1 or PD-L1, will directly kill cancer cells. So in this case, the killing cell is a T cell not a macrophage. Despite that we have proven that in this setting we compare M2 macrophages that they are immunosuppressive to M1 macrophages that have more anticancer, which is another mechanism that plays part in combination with checkpoint moderators. We have run 2 studies in combination with KEYTRUDA. One that was in second line evorpacept plus KEYTRUDA. Only 10 patients had a response rate of 40% and overall survival rate of 12-month of 80%. That compares very well with the results of KEYNOTE-40. Also in the first line head and neck, we combined evorpacept with KEYTRUDA and chemotherapy. This was a small study. In this case, the response rate is similar to the KEYNOTE-48 benchmark. But also, again, the overall survival is much longer, 12 months, than the KEYNOTE-48. With this positive data, we're now moving or actually in the middle of 2 Phase II studies as both of them randomized, in first line head and neck. In one, we are combining evorpacept with KEYTRUDA, and the other one, evorpacept and KEYTRUDA plus chemotherapy. There's the 2 labels of KEYTRUDA in one case in the doublet is for patients that have PD-1 scored higher than 1. While in the KEYTRUDA chemotherapy is PD-1-agnostic. So any patient can be enrolled in this study. The endpoint is 12-month overall survival and response rate. Moving to hematological malignancies. On Slide #19, you have the mechanism of evorpacept, mechanism in combination with azacitidine. Azacitidine when it gets internalized into cancer cells, regulates calreticulin. Calreticulin is another eat-me signal that interacts with LRP. That's a receptor of macrophages and provides the required positive signal. In this setting, blocking CD47 enables macrophages to eat cancer cells. On Slide #20, you can see the study that we are running now. We have a dose escalation safety study, where we are both enrolling newly diagnosed MDS patients and relapsed/refractory MDS patients. We're starting at 20 mg per kg every other week, 30 mg per kg every other week, and then we move to 60 mg per kg every 4 weeks to be able to conveniently combine with azacitidine that is dosed once a month. We have proven safety up to 60 mg per kg already, and now we have moved into dose expansion. You can see on the right, our patient characteristics, that more than half of them are refractory, where azacitidine is not expected to have much effect. And many of them has very complex cytogenetic status. For example, we have a unusually high percentage of TP53 patients that have a lower prognostic. On Slide #21, you can see the data of the part of the patients in the dose escalation safety study. It consists of a very small number of patients, but we have a lot of the TP53 patients, 2 CRs of the 5, and that compares very well with the data that has been shown for magrolimab, the Gilead molecule. And also, it's clearly higher than the expected response rate of azacitidine that is more -- 17% to 20%. In the refractory setting, where azacitidine is not expected to produce much of an effect and no other CD47 molecule has shown any data, you can see that we have 5 patients out of 9 with a marrow CR. That is very interesting to us and provides a clear proof of activity of evorpacept in this setting. Again, with this promising data, we are now running a dose expansion -- dose optimization study following FDA guidelines of trying to find the dose before moving to randomized studies in where we're comparing 40 mgs per kg every 4 weeks or 60 mg per kg every 4 weeks, both doses, we believe, are efficacious. And from there, we're going to move into a randomized study. We're going to compare evorpacept and azacitidine versus azacitidine and the regulatory endpoint is complete response. I think the activity of evorpacept in the clinic is better seen when you actually compare all the studies together. So on Slide #24, you have the data of solid tumor for evorpacept compared with regulatory benchmark on those indications. In the left, you see the second HER2-positive gastric cancer. The combination of evorpacept, Herceptin, CYRAMZA and paclitaxel has a response rate of 72%, where the benchmark is 28%. And more importantly, the overall survival of 12 months of close to 80% doubles of HER2 is seen as a benchmark of 40%. And also, this data is superior to the data shown for the -- in HER2 in the resting study, where response rate is 40% and overall survival 12 months is about 50%. In head and neck, first line, in combination with KEYTRUDA, 5FU and platinum, the response rate is very similar. But that is the same that we're seeing in KEYNOTE-48 comparing KEYTRUDA chemotherapy versus the contralateral in KEYNOTE-48, where the response rate was identical as well. And KEYTRUDA was approved based on overall survival at 12 months. If we focus overall survival at 12 months, we see that in this very small study, overall survival for the evorpacept combination is 88% versus 53% in the KEYNOTE-48 KEYTRUDA arm. Same situation in second line. But here in absence of chemotherapy, we see some improvement in the response rate with a 40% versus the benchmark of 15% in KEYNOTE-40. But again, more importantly, we see a very significant improvement in OS rate at 12 months, 80% versus 37%, and this is with 3 years follow-up. In Slide #26 -- 25, we see the same for hematological malignancies now comparing with magrolimab, the CD47 molecule of Gilead. As I said before, in the TP53 mutants, magrolimab had 2 CRs out of 4, we have 2 out of 5, so very similar. And we have a much better safety profile. As I mentioned before, magrolimab did not publish any data in refractory patients where we have very interesting activity. In NHL, in combination with rituximab, also the response rates and the CR rates are very similar with magrolimab and evorpacept, but again, in our case, with a much better safety profile with no -- nearly no dose-dependent cytopenias in our case and very high rates of cytopenias seen in the case of magrolimab like anemia, thrombocytopenia and neutropenia. Based on the data that I just showed you, we believe that these indications are just the tip of the iceberg that if it ends that we prove one of these mechanisms in a randomized study, we'll be able to move into other combinations with other indications where those mechanisms are relevant. On Slide #27, you can see that 2021 was a very busy year for us with a lot of studies initiated a readout and that continues in 2022 where we have interesting readouts, for example, the MDS dose optimization study and also initiation of many other studies and also data that can come from our recent collaboration. 2023 is when things will get really interesting with the readout of our gastric trial, randomized, and also the AML study and the IND filing for our second compound, the SIRPalpha CPG conjugate. And in 2024, we will have also the results for the randomized study in head and neck. With that, thank you for your attention. And I want to introduce Dr. Sophia Randolph, our Chief Medical Officer, to answer any questions. Daniel?

Thank you for the presentation. And as a reminder, if you want to ask a question, please use the blue Submit a Question box and I'll be able to ask questions to the portal. There's a couple of questions. Maybe to start with one of them. While acknowledging the rationale for a combination approach, given the mechanism of action with different modalities, how would investors go about assessing the true value proposition of evorpacept as a monotherapy? What does it bring to the different treatment modalities that you're putting it into?

Yes. Evorpacept was designed to be used in combination. So to actually have a large therapeutic window, we sacrificed the monotherapy part of it. Therefore, in every single study, we have to be combining it with something that specifically targets cancer cells. How we evaluate that. Now how the investors can see the activity of evorpacept in this context, you have to look at all the studies, one by one or in context before we have randomized data. And of course, the real answer will come in the randomized studies. The [indiscernible] drug developer is that the signs of activity that we have in 5 indications give me a very good confidence that evorpacept is adding to these different studies. But maybe Sophia?

Yes. No, just as you've said, I think once you get into the clinical studies, it's really comparing with historic controls and seeing the improvement in clinical activity with the tolerability compared to standards of care, seeing that sort of reproducibly across different combinations, different tumor types does give us the confidence that the drug is actually doing something positive in these different indications. But true proof of concept will come in the randomized portion of the development and that's the portion that we've entered into now.

When looking at the gastric cancer landscape and specifically the HER2 plus gastric cancer in the second line or greater, with KEYTRUDA as part of the first line of standard of care recently introduced, do you expect patients to respond differently in second line with evorpacept combination when you start to do -- start to doing this randomized Phase II as opposed to earlier Phase I studies when patients were not seeing that KEYTRUDA in the first line?

So yes, it's a great question. Certainly, there has been a sea change in the first line setting, which is great for patients with the introduction of pembrolizumab in combination with trastuzumab, a fluoropyrimidine and a platinum for HER2-positive disease. The nice thing about the evorpacept studies that were done in the second line or greater setting is we actually did enroll patients who had come off some of the clinical trials that were underway exploring the pembrolizumab activity even before it was formally approved. So within our second line or greater setting, when we were looking at evorpacept plus trastuzumab doublet, approximately half of the patients had actually already seen a checkpoint inhibitor in the first line setting in combination with trastuzumab. And we were able to see partial responders in that population. Similarly, when we look at our 4-drug regimen, so ALX or evorpacept plus tras, CYRAMZA, paclitaxel or Herceptin, CYRAMZA, paclitaxel, there too, although we only had 2 patients who had seen prior checkpoint inhibitor, 1 of those patients also achieved a partial response. So with the numbers that we've evaluated so far, and it makes sense from a mechanistic standpoint, we don't believe seeing checkpoint inhibition an inhibitor of the first-line study should impact evorpacept activity in the second line or greater setting.

Great. And then maybe continuing within this setting, what do you plan to see in the Phase II ASPEN-06 study that would convince you -- that would drive the advancement of the study into Phase III?

So the study -- the Phase II portion of the study for ASPEN-06, as Jaume explained earlier, is really aimed at looking at the contribution of evorpacept to a tras-containing regimen because it is a 4-drug regimen, and we want to be able to sort of elicit what exactly is the evorpacept contribution to the 3-drug backbone. So for Phase II, we'll be looking at evorpacept plus Herceptin, CYRAMZA, pac versus Herceptin, CYRAMZA, pac. There, we're looking for at least a 10% difference between the 2 arms. And then also, we have -- the study is powered to look at the comparisons of each arm to historical control ram pac or CYRAMZA pac, which is the true global standard of care. So those are sort of how the study is being powered, so looking for a magnitude difference of 10% between both arms, and then ideally looking at an improvement from a historical rate of approximately 30% for -- or 28% or 30% for CYRAMZA, paclitaxel, looking to get to about 50% response rate on the test arms. Given what we've seen in the ASPEN-01 first in-human study in that cohort, where we've seen a 72% response rate, I think we're very encouraged that, that Phase II will hopefully read out positive.

Okay. Great. In MDS, we have seen great efficacy in Phase I that necessarily did not translate to separation in randomized studies as recently as last year. And so I guess what are you trying to ensure that when you go to randomized studies, that, that efficacy will be able to continue to be interpreted as coming from evorpacept combination as you go from single study to a randomized controlled arm studies in MDS.

Yes. So for MDS, it's been interesting. For the longest time, azacitidine was the only real -- or a hypomethylating agent was the only real active agents in this particular space. So those original studies, which quote in higher-risk MDS patients quote a CR rate of about 17%. And those studies were done probably over a decade now at this point. Recent studies, as you allude to, the azacitidine monotherapy activity has been bouncing around a little bit kind of in the low 20s. And then most recently with the PANTHER study, there was a aza monotherapy rate. That was actually very high. It was around 30% and seems to be a bit of an outlier compared to other randomized studies that have been done recently. A meta-analysis done in 2020 shows the azacitidine monotherapy rate to be netting out around 20%. And I think KOLs at least that we've spoken to feel that, that's a very reasonable benchmark. But it does speak to the importance of randomization. And so despite seeing the activity, and again, for us, it's very early days. The data that we presented at ASH was for TP53, again, 5 patients who were response-evaluable at that point. But it does speak to the importance of randomization in -- when looking at proof of concept. So in the Phase I setting, it's appropriate to look at our activity versus historic control. We benchmarked azacitidine around 17% to 20%. But then as we go into the Phase II setting, as Jaume described, here we'll be looking ultimately comparing evorpacept plus aza versus aza alone. And it just has to do with the heterogeneity of the population as well as, I think, just the duration and time from when those original studies with azacitidine were done. So we'll be looking for an improvement ultimately over an azacitidine backbone where the assumption of the aza is about a 20% response rate.

HER2 breast cancer is a substantial opportunity. And so with the recent collaboration with Zymeworks, I'm just wondering, what are you looking for in the -- first of all, I guess, what was the rationale into forming that collaboration, particularly going to a study that also looks not only at the high expression of HER2 but also low HER2 expression? And what does evorpacept bring to the table that would allow you to be able to bring a differential efficacy measure?

Yes. As you know, most -- Herceptin mostly works for blockade of the signaling. And so the mine activity of -- in the clinic of Herceptin is HER2 signaling blockade. It doesn't have much of a ADCP activity. We picked Zymeworks molecule because there is a bispecific molecule that has 1 epitope that binds the same epitope of Herceptin and the other arm binds the same epitope of trastuzumab, which provide a much higher density of Fcs on the cancer cell. So providing that positive signal that we can enhance with the blockade of CD47. So that would make sense in the low HER2 setting when you have less receptors, but now with this molecule, you can have twice the number of antibodies bound to those receptors, and it could be enhanced by the blockade of CD47. So that was the reason we picked that molecule to combine with evorpacept, but maybe, Sophia, you have any clinical point of view?

Yes. No, I think -- and just from a bucket in terms of a mechanism of action, I think based on our gastric data from our first in-human as well, even as our non-Hodgkin's where we're combining with anticancer targeted agents such as rituximab, this mechanism of action where we're enhancing that ADCP activity lends to other indications. So whether it be other indications where HER2, for example, is targeted or even other solid tumor indications where there may be other anticancer targeted antibodies. These are some of the additional indications that we're interested in expanding into.

Yes, for example, colorectal, right. So we have an IST combined with colorectal plus cetuximab and PD-1 and [indiscernible]

Maybe bigger picture, as you think about future collaborations, what are some of the underlying drivers, both in terms of unlocking commercial value for evorpacept, but also in terms of the right mechanism of action that works -- as you think of -- so yes, both in terms of unlocking the opportunity as also mechanistically works the best, how are you thinking about it actually when you think about future collaborations?

Yes. So we're trying to zeroing the mechanisms and then expanding from there. So I think the mechanism in combination with anticancer antibodies in our mind is very well validated. So then that's why we already have a collaboration with Zymeworks with HER2 positive. And we're going to be cetuximab, as I just mentioned, in the case of colorectal and we're going to be actually trying to do our collaborations potentially in antibody drug conjugates as well, where I think it makes a lot of sense to combine. The same with the checkpoints. We have now head and neck and colorectal study just mentioned, but we're going to be actually looking for other indications where checkpoints are used. And in the case of small molecules like azacitidine, we have already 2 studies, our AML and MDS. But from there, we will see. So, Sophia?

Yes. So exactly that. So when it comes to future studies, I think ADCs and solid tumor are one that are of a special interest to us. And then exploring a little bit more the combinations with the small molecules, we certainly not only with azacitidine, but even across our other studies where we've been combining with chemotherapy, where we're seeing activity or potentially even radiation that might be another area. But I think for the short term, it's looking at these 3 mechanisms: anticancer antibodies, checkpoints and then small molecules. And that's how we're moving it forward.

Great. Maybe we can bring that to wrap. Again, thank you very much, Jaume and Sophia, for your presentation and also for your thoughtful answers. And thank you for our investors for listening in. Thank you.

Thank you, Daniel.

Thank you.