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

All right. Perfect. So welcome, everyone, to the 30th Annual Healthcare Conference of Crédit Suisse. I'm Tiago Fauth. I'm a biotech analyst here at Crédit Suisse. We're joined today by ALX Oncology for a fireside chat. Feel free to e-mail me any questions that you guys want to try to -- I'll try to work those in. We're joined today by Jaume Pons and Sophia Randolph from ALX Oncology. And basically, we can always start with some brief remarks on some of the highlights of 2021 for ALX Oncology so far. We're going to talk a little bit about the competitive landscape and then dive in to some recent data updates. But Jaume, you can take it away for a brief intro and I'll -- we'll follow from that.

Yes. So I think that this year has been a good year for us. We have a lot of data coming out, and we have started many Phase II and other trials that we've been in for years. We're very excited about our data so far, both in hematological and solid indications. I think that our initial hypothesis that we made when we started the company, so far, is meeting the expectations.

Got it. And to that point, I guess, we can now start questions about the theoretical differentiation, right? So CD47 became a relatively crowded, at least perceived to be a relatively crowded, competitive landscape. What's the biggest opportunity for differentiation for your assets specifically? How does it differentiate from other CD47 programs out there? And how they may -- that may actually show up in clinical data?

Yes. So in my opinion, CD47 is a mechanism that is going to show its potential in the combination setting. The main differentiation between us and everybody else is that we designed our compound to be used in combination. And therefore, by inactivating [ that C ], we have a much better safety profile, we have the ability to dose higher. And in the clinic, we have shown that higher dosing translate to more efficacy. All the other companies design compounds that try to have some single-agent activity. This single-agent activity has been extremely modest, and then it was -- always, all of them being developed with a compound, mostly in combination. So I think our advantage is that the compound was designed to be used in combinations. We'll be able to go to earlier lines of therapy in combination with chemotherapy, for example, because we are not going to have overlapping toxicity. And at the same time, we are able to move in combination with multiple agents that will have minimum safety concerns. Because our track is [ better curated ] , these combinations will be possible. So in a few words, we designed the compound to be used in combination. It's been developed in combination. Everybody else designed their compounds to show single activity. When they are being developed in combination, they carry some baggage in terms of safety.

Got it. And what are some of the key features from a safety perspective then that in clinic has kind of differentiated you guys relative to the competition? And to be fair, like -- so sometimes when you speak to physicians, they seem to be comfortable with the overall profile so far of most CD47 agents. So when does that potential safety differentiation actually plays out? Does it necessarily require more combination trials and moving earlier in the treatment paradigm? How can that play out for you guys?

Yes. Well, let me start that, and then, Sophia, you can pitch in as needed. So the way I see it is that so far, other data from other CD47 molecules has been mostly hematological settings, where transfusions are usually part of the treatment, part -- like MDS, AML, right? So the main tox of CD47 molecules is a hematological tox. So in those settings where transfusions are used very frequently, this hematological tox can be hidden by the transfusions that are given to the patients. When you move to other indications where patients are not transfused frequently is when the anemia, thrombocytopenia, neutropenia that other CD47 molecules show in the clinic are more relevant -- and especially if you want to move in the first line or earlier lines, where patients are not using transfusion all the time. So Sophia?

Yes. No, exactly. And it's important when you think about the clinical development strategies for a compound because, ultimately, you do want to get into the earlier lines of therapy where often chemotherapy or other cytotoxic agents are commonly used. So with evorpacept's unique safety profile, it allows for those types of combinations to be well tolerated. We don't -- to date, we haven't been adding any toxicity to the backbone of whatever our combination partner is. So it allows us to escalate to higher exposure levels as well as bringing up into earlier lines of therapy, where we've been able to combine with multi-agent chemotherapy regimens and other apoptotic-type cancer regimens.

Got it. No, that's perfect. And I mean solid tumors is definitely one of the key drivers for the story here. But perhaps before we go into that discussion, maybe just a couple -- a few minutes talking about hematologic malignancies, right? So you recently had an ASH abstract in high-risk MDS. There was some noise relative to the abstract and the press release and how exactly to contextualize those data. And we're talking about 5 patients' worth of data, right, to be frank. So a lot of extrapolations from a relatively small sample size. But perhaps what is the best way to look at those data in your opinion? And how does it actually stack up versus prior therapies in the CD47 space? And I'll ask a couple of follow-ups on that.

Okay.

Yes. Yes, you're completely right. It's 5 patients, right? So I will let Sophia go over it, but you have to take it with 5 patients' data. If you look at these 5 patients, in my opinion, the data is looking good. There's not any reason to believe that it's going to be any worse than magrolimab, but no. So Sophia?

Yes. So definitely, between both the abstract and the press release, what we were able to present with that abstract data at ASH over the past couple of days has been 2 pots of data. And so the first pot of data from a response-evaluable perspective was looking at 5 patients who were newly diagnosed with high-risk MDS and then another 5 patients who are in the relapsed/refractory setting. And the data was really encouraging. And essentially, within those first 5 response-evaluable patients, all of whom just happened to be TP53 mutated disease. Those -- of those 5 patients, we have 2 who have achieved complete responses. So it is 5 patients, it's about 40%, which is on par with the 40% that's been described previously with magrolimab as well in their Phase Ib. Certainly encouraging compared to the backbone, which would be azacitidine alone. Our combination was evorpacept plus azacitidine, where you would expect somewhere around 17% to 20% objective response rate. On the relapsed/refractory side, there, also very interesting. We saw 2 patients who presented with a marrow CR. Anything in the relapsed/refractory setting is considered exciting. I don't think magrolimab has actually published -- at least I'm not aware of any data that they published in the -- looking at magrolimab plus azacitidine in relapsed/refractory. They do have a monotherapy data out there where they saw, I think, 1 marrow CR out of 10. So between those 2 pots of data, we're encouraged. The drug has been well tolerated. As was stated in the abstract, the goal was to get up to 60 mg per kg every 4 weeks, which we were able to do. And remember, given the size of the molecule, that's equivalent to 120 milligrams of an antibody. So safety-wise, not adding on to the -- any toxicity profile beyond that of azacitidine alone or the disease itself. And then we'll be presenting more data at ASH in about 20 patients.

Okay. So -- and with 20 evaluable patients, is that mix of relapsed/refractory and newly diagnosed? Or...

Yes. It will be a mix of both. And the 20 -- approximately 20 was for safety. Those who are response-evaluable will be a slightly smaller group, but it -- yes, yes.

Got it. And you did allude to at least the TP53 mutation. From baseline characteristics, is there anything else that we should take into account when trying to comp this versus magrolimab? How important is actually that specific mutation in terms of characterizing prognosis and in context of the overall data?

Yes. So TP53 in general is considered a poor prognostic characteristic. The actual response rates in TP53 -- patients who have TP53 mutations for newly diagnosed MDS, response rates are not that far off from the wild-type population. It's more about how do they go on to progress and their overall survivals where they have a worse prognosis than those who do not. So even with the magrolimab's MDS data in patients who had TP53 mutations, they've only presented really data on 4 patients from their earlier data. And of those 4, 2 had, had a CR. So it's in line with, I think, some of the TP53 data that's out there from other competitor agents, where TP53 response rates, wild-type response rates are fairly similar. It's really more about how do those patients perform as time goes on.

Got it. And I guess as we have fewer benchmarks for the relapsed/refractory patient population, I was wondering if you guys could outline -- as you mentioned, almost any activity would probably be perceived as positive. But what's the hurdle rate for that market? And in all honesty, is that a -- how big of a market could that even be given how rapidly progressive the disease can get at that stage?

I mean it's a very aggressive patient population. So really the -- if anything, the medical need is most there, right? That's where patients really don't have much in the way of treatment options. So the bar for objective response rate in that population is extremely low. But I think first, the point is can we see a clinical improvement, and that's always the first step. And so that's the data that we'll be presenting at ASH and then worry about the market size after that. This is whether or not we actually see some improvement, are we providing clinical benefit in that population.

Got it. No, that's absolutely fair. So perhaps just in terms of trying to understand expectations for the fuller data set with about 20-something patients. You did show some deepening of responses relative to the abstract, and I'm assuming that was mostly just a function of the cutoff date before you submitted for the newly diagnosed patients. Is it fair to assume that the actual fuller presentation with more patients will have a more mature data set? Or is there a chance for that data to be in a continuous spectrum or you can see better responses with a longer follow-up in a subset of those patients? Just trying to understand how mature do you think that data set will be at ASH.

Yes. So the data that was in the abstract and the press release are all from the same...

Same type of patient, yes.

So the difference between the 2 is just, as you know, for -- as you're evaluating patients with this disease, there are a lot of components that go into the final overall response assessment for a patient. So at the time of the abstract, not all of that data had come back yet. And -- but with the press release, we were able to update with the complete response data set for each of those patients. But they are all as of the same assessment date. For ASH poster, we will have a more mature data set on all of those patients. So we would expect some additional data coming from both pots of data that's new.

Got it. No, that's perfect. So I guess we can pivot probably to solid tumors. So just bigger picture, I mean different CD47 companies have taken a different strategy towards solid tumors, right? So not necessarily a whole lot of overlapping indications relative to hematologic malignancies, where everyone is kind of pursuing a similar playbook to some extent. So what drove your choice of indications, specifically relative to perhaps what Forty Seven was doing or some other companies in the future might try to pursue?

Yes. So when we're looking at initially the strategy to develop our compound, we had to think about indications where there were drugs that were approved but they didn't work very well. So actually, we had a window to see improvement over those drugs. Also, we were looking about the relative CD47 expression, so incredibly high and also presence of macrophages in the tumor. So we found that gastric and head and neck were 2 indications. We were looking in terms of mechanisms as well. So for each one of these indications, we're testing a different mechanism. In the case of gastric cancer, we're testing the combination with anticancer antibodies that are targeting a cancer antigen, for example, Herceptin in the case of gastric. In the case of head and neck, we are testing the combination with checkpoint modulators. So that -- there's the other aspect of the election of the 2 indications. In one, we're testing one mechanism; in the other one, another mechanism.

Got it. So I'm assuming that also kind of addresses the question that you've probably gotten a few times, which is the mechanistic rationale to pursue solid tumors, given that you haven't really seen a lot of success previously with other agents. So perhaps it's about just finding the right indication with the right combo and mix of prophagocytic signals and checkpoint inhibitor synergy. Is that the right way to think about it?

That's part of it. As usual in oncology, right, finding the right tumor for the right drug is part of the story. The other aspect is that because we have a very good safety profile, we have been able to dose very high in those indications. Also, as Sophia told you, that our molecular weight is half of an antibody. So potentially, we have a better tumor penetration in a, sorry, tumor setting. Also, our affinity is the highest one. We have a picomolar molecule that are actually, in the limiting amounts of molecule that can get inside a tumor, we may have enough affinity to totally block the mechanism where others may not. So I think all these 3 plus the chosen combination partner [ may be perfect ] in that we have been successful where others have not.

Got it. Understood. And understandably so, there has been a lot of focus on the upcoming head and neck, especially first line -- the frontline data. That's mostly because you've had some pretty interesting results but a small sample size as well, so 4 patients' worth of data in the frontline, though far exceeding what would have been expected from benchmark at that comparable -- fairly comparable benchmark in this case. So for the update, as you think about growing that sample size, what should be expectations in terms of number of patients or reversion to the mean? Is it most important to look at response rates or perhaps duration of response, overall survival? What are going to be kind of the same drivers and the best way to look at that frontline data?

Sure. For the head and neck population, first off, for SITC, we'll be presenting data on about 13 patients, so we've -- who have been treated with evorpacept, with pembrolizumab, 5-FU and cisplatin. So that will be the new data that we present and then we'll be -- in the first-line setting. And then we'll be also presenting follow-up survival data. We'll get to that survival question right after this. We'll be presenting survival data on the original cohort of 10 patients who are in the second line setting for head and neck but who were checkpoint-naive. And those patients received evorpacept plus pembrolizumab, that doublet. So those are the 2 pots of data that we'll be presenting. In terms of the newer cohort, the chemotherapy combination cohort, so with that particular group, in terms of benchmarks, you're looking at KEYNOTE-048 as a relevant benchmark, where you had patients receiving pembrolizumab plus 5-FU and cisplatin versus the extreme regimen, which was cetuximab plus 5-FU and cisplatin. And one thing that is important to note from there, anything that you can see in terms of objective response rate is significant. But in terms of approval of pembro plus 5-FU and cisplatin, it was really based off of the survival data. And it has to do with immuno-oncology agents and their impact on longer time points, such as overall survival or later line in terms of PFS. So as we move forward with the data that we show at SITC, anything that we can see above benchmark in terms of objective response rate is important. And then, in addition, looking at survival-based end points as surrogates for clinical benefit in this population is important as well. Similarly, for the second-line setting, again, any improvement in objective response rate is definitely important. But then it's also important to look at the longer time points that are the survival-based endpoints. And we'll be presenting all of that data at SITC.

Got it. And that's -- this is perhaps a fairly subjective question, but whenever you're running a trial without a control arm, what's the magnitude of improvement that actually gives you confidence that, that could translate into a real signal in a controlled fashion? So if you're talking...

Yes, yes.

Yes, sorry. If you're talking about a progression-free survival of 4.9, right, I think, for benchmark in frontline, are we talking about months? Are we talking about a 20% improvement? What would be interesting enough that would warrant further development? And I guess if you can frame what would be some of those expectations.

Sure, sure. I think in terms of -- for head and neck, really, again, as you go back to the KEYNOTE-048, the endpoints that were particularly predictive of response were actually the overall survival metrics and overall survival rates. And there is where you started to see those Kaplan-Meier plots split apart between the test and the control arm. So they're looking at anything -- 30% to 50% improvement would be, again, the gold standard of overall survival or overall survival-based metrics like OS at 12 months. That would be definitely, I think, impactful. It's -- as a single-arm study, you're always limited because you are comparing to historical controls. And over the course of time, patients are treated differently in the prior lines of therapy. And so sometimes that can impact how you can interpret your results. So ultimately, you want to get to randomization as, obviously, the key. And that's what we started now with our ASPEN-03 and ASPEN-04, both studies in the first-line setting, evorpacept plus pembro/chemo versus pembro/chemo. And then in ASPEN-04 -- or that was ASPEN-04; ASPEN-03, evorpacept plus pembro versus pembro. And so those 2 Phase II studies are actively accruing now.

Got it. Perfect. And perhaps just a quick update on gastric and kind of what's the status for that program overall would be helpful.

So for the gastric program, again, as Jaume mentioned earlier, there, we're testing a slightly different mechanism of action. We're looking at evorpacept plus anticancer specific antibody, so looking at its combination with trastuzumab. At SITC, we will also be updating the data that we showed at ESMO GI, where we were looking at that doublet, ALX plus trastuzumab on top of the current global regulatory standard of care, which is ramucirumab/paclitaxel. So at SITC, we'll be presenting updated data on those 18 patients who received ALX plus tras/ram/pac. And it will have some longer duration of exposures and therapy and responses that we'll be updating there. But the data even from ESMO GI, where we saw the 72% objective response rate, where your benchmark for ram/pac is about 28% and a median overall survival that hadn't been reached and a doubling -- more than doubling of median PFS, that data was extremely encouraging in that expansion cohort. And it's the reason for us working towards our ASPEN-06, our randomized Phase II/III, where we'll be looking at that same quadruplet versus -- in this case, versus tras/ram/pac. So that's coming up. The study is up on ct.gov, and we're working on activating the sites.

No, that's perfect. I guess anything else within solid tumors that we haven't covered that might be worth highlighting to investors at this moment? Otherwise, I'll just jump in from a business development perspective and a couple of questions there.

No, I don't think so. All that and we'll continue running our study with Zymeworks in [ breast cancer ].

That's correct, yes. No, perfect. So from a business value perspective, I mean, here, we get a lot of questions on M&A potential in the CD47 space, right? You have Forty Seven. You have Trillium as well. So given that there certainly seems to be a lot of interest in -- from a strategic perspective, from large biopharmaceutical companies, how do you think ALX kind of stacks up versus other unencumbered CD47 assets that are still potentially available, I guess, for a potential transaction? And how should investors think about that potential?

Yes. As I said before, in terms of design of the molecule and the clinical development path, we're very differentiated because we have a molecule that was designed to be used in combination and is being developed in combination. And with a strong emphasis in solid tumors as well, right? So in that, we are very unique. The other 2 acquisitions were mostly focused in hematological malignancies. So I think that this is a very big space for us. Actually, the market for solid tumors is larger. So I think that as we progress in our development, there's potential for negotiations. But at this moment, we are completely focused in our job of actually developing evorpacept.

Fair enough. Fair enough. And you also had a couple of -- you had a recent acquisition, ScalmiBio. So what's actually behind the rationale of ALXO's strategy in terms of doing a couple of partnerships here, a tuck-in acquisition there? Where do you think that -- how do you think that complements your current pipeline? And what's the overarching strategy here?

Yes. So we are in the process of building a pipeline. So far, 100% of our focus was in evorpacept. Then what would be next is a collaboration with Tallac to have a SIRPalpha CpG compound, which I'm very excited about. And we should be filing IND maybe end of next year or so. And then ADCs is something that always had a lot of interest. And I think that at this moment that we have proven that we have increased activity of anticancer antibodies and as well potentially chemotherapy. ADCs make a very good partner for ALX148. There are not that many interesting approved ADCs in the solid tumor setting. So we decided that it was time for us to have our own set of ADCs as well. Obviously, at this moment, we're helping other compounds to get a bigger market. Why don't we do it with our own compounds? So that's the idea of starting an ADC effort for developing compounds that will stand on themselves, so they will not be designed just to be combined with ALX148. So they will have a life of their own and indications of their own but also could be enhanced by evorpacept.

Got it. Got it. No, I think that makes sense. Great. Honestly, that kind of runs through most of the questions that I've got and that I had prepared. Is there anything else that you think investors are kind of missing around the story? Stock has been fairly volatile around data updates, but the question is always kind of the same, which is just how do you compare against the rest of the landscape, right? And without any head-to-head data, I think it's kind of difficult to make an extrapolation. So perhaps one final for me that is related to the competitive landscape. But you do have an upcoming potentially pivotal readout for magrolimab in MDS. So I don't know, based on what's the clinical experience in MDS for your own molecule, what are the potential read-throughs that you could have from that readout to you and perhaps to the CD47 class more broadly?

Yes. In my opinion, [ if that's all right ], magrolimab in MDS and AML [ compete ] with the mechanism. I think that the mechanism is proven in the clinic. And that is what actually is most important for me of the work of Forty Seven, now Gilead. I personally don't think that Gilead will get approval with a single-arm study in MDS. And I think they will have to finish a history, which actually gives us an opportunity to be able to catch up, if that is the case.

Got it. And is that mostly from a sample size perspective or data? Like what's -- what informs that opinion?

My opinion that it will not be is because I think -- that's obviously a personal opinion. But I think that in a combination with something like azacitidine that is active -- and there recently have been a few studies where the benchmark for azacitidine has been bouncing.

Got it. There's been a lot of uncertainty and unpredictability. Okay. That's a fair point. Now that's...

It was a very heterogeneous population. So you can get differences depending on who's actually enrolling. And it's one of those diseases that's very complex in terms of the baseline characteristics of patients.

That's fair. And it's not like you don't necessarily have a lot of precedent comps as well to inform you on that, right? So got it and understood. Point well taken. Great. No, I think that's it for me. I don't know if you guys have any final remarks, but extremely helpful. So I do appreciate you spending the time with us. And hopefully, it's going to be a productive conference for you guys as well. But thanks a lot for your participation. Appreciate it.

Thank you. Thank you.

Yes. Thank you.

Bye.