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

Good morning, and welcome to the ALX Oncology Conference Call. [Operator Instructions] I will turn the call over to Jeanne Jew, Chief Business Officer of ALX Oncology. Jeanne.

Good morning, and thank you for dialing in today's conference call. I'm Jeanne Jew, Chief Business Officer of ALX Oncology. Today's call will focus on clinical data of Evorpacept, a CD47 myeloid checkpoint inhibitor and specifically, data from our ASPEN-01 Phase Ib trial in gastric cancer and head and neck cancer that will be presented at the Society for Immunotherapy of Cancer or SITC on November 13. A slide deck for this webcast is available on our website in the Investors Events section. Dr. Sophia Randolph, our Chief Medical Officer, will discuss the updated Phase Ib gastric cancer combination data. We are also privileged to be joined today by Dr. Kevin Harrington, Professor of Biological Cancer Therapies and Head of the Division of radiotherapy and imaging at the Institute of Cancer Research, who will discuss the clinical data in head and neck cancer in combination with pembrolizumab and chemotherapy. At the end of our remarks, we will open the line for questions. Dr. Jaume Pons, CEO, will also be available during the question-and-answer session. Before we begin; we would like to remind you that today's presentation will include forward-looking statements based on our current expectations and beliefs. Such statements represent our judgment as of today. And involve substantial risks and uncertainties that may cause actual results to differ materially from the results discussed in the forward-looking statements. With that, I would like to introduce Dr. Sophia Randolph. Sophia?

Thanks, Jeanne. Today, I will be highlighting data on evorpacept from the gastric cancer portion of the ASPEN-01 Phase I trial from poster 498 that will be presented at SITC 2021 this year. On Slide #1, you can see that gastric and gastroesophageal cancers are one of the more common cancers worldwide with greatest incidents in Asia and poor survival in the advanced setting. By way of background on Slide #2, standard first-line therapy for patients with advanced HER2-positive gastric cancer is trastuzumab plus chemotherapy, most commonly a doublet of a fluoropyrimidine and a platinum. In the second-line setting, the combination of ramucirumab and paclitaxel or RAM + PAC is associated with the best outcome and is considered the global regulatory standard of care. Trastuzumab-deruxtecan was evaluated in the third-line setting in patients with HER2-positive gastric cancer in the destiny gastric 01 study. This randomized Phase II study was positive based on objective response rate and overall survival. And has been approved in the U.S. in the second line and later setting and is a standard of care for patients as third line therapy in Japan. However, trastuzumab-deruxtecan challenging safety profile may make its use in earlier lines of therapy challenging. Preliminary report at ASCO this past year of its use in patients with HER2-positive gastric cancer in a single-arm Phase II study conducted in the United States and in the EU in the second line setting had a similar objective response rate and PFS, but a shorter duration of response compared with that in DESTINY-01. Another second-line regimen, trastuzumab, ramucirumab/paclitaxel was evaluated in a single arm, predominantly single institution Phase II study where an initial 52% objective response rate and median OS of just over a year were reported. Despite these recent advances, novel and tolerable regimens are still needed for the treatment of this aggressive cancer. As noted in Slide 3, tumors upregulate CD47 as a way to evade the innate immune response. EvorpaceptEvorescept is a high-affinity fusion protein that blocks the CD47 SIRPalpha myeloid checkpoint interaction. Because the molecule has an inactive Fc it spares normal cells from CD47 targeted antibody-dependent cellular phagocytosis or ADCP activity. Evorpacept is designed to enhance the activity of anticancer targeted antibodies and checkpoint inhibitors with minimal hematologic toxicity. By inhibiting the myeloid checkpoint, evorpacept in combination with trastuzumab with its active Fc domain maximizes the macrophages HER2-targeted ADCP activity. Evorpacept also directly activates dendritic cells, enhancing their cross-priming of T cells and so T cell activity in combination with checkpoint inhibitors. Now on behalf of the ASPEN-01 investigators in Slide #4, I will present the Phase I data as of September 1, 2021, summarizing evorpacept for patients with advanced HER2-positive gastric or gastroesophageal cancer presented at the annual meeting of FITC 2021. As you can see on Slide 5, 18 patients with predominantly second line HER2-positive gastric cancer received ALX148 or evorpacept at 10 or 15 mgs per kg every week in combination with trastuzumab, ramucirumab and paclitaxel. The study enrolled patients from the United States and South Korea, with more Asian subjects enrolled into this cohort. The majority of patients had distant visceral metastases, an ECOG score of 1, and all but 1 patient had disease that had progressed upon prior anti-HER2 therapy. On Slide #6, evorpacept was well tolerated in combination, the most common evorpacept-related adverse events in combination with trastuzumab, ramucirumab and paclitaxel were low-grade rash, urticaria and diarrhea. There were no dose-limiting toxicities or study deaths or evorpacept-related serious adverse events. The maximum evorpacept administered dose in combination was 15 mg per kg weekly. Here in the table on Slide 7, we see that in 18 patients with second-line gastric cancer a 72% objective response rate was seen with a median duration of response of 14.8 months. I'd like to remind you that the objective response rate with ram-pac in the RAINBOW study was 28% with a duration of response of 4.4 months. And in the DESTINY-01 gastric study, the trastuzumab, deruxtecan ORR was 41% with a duration of response of 11.3 months. The longer-reaching impact of evorpacept is especially seen in the preliminary median overall survival of 17 months which compares favorably with the 9.6 months median OS seen with ram-pac in rainbow as well as with trastuzumab deruxtecan's 12.5 months median OS in DESTINY-01. The waterfall and spider plots show the magnitude and deepening of response over time, in some cases, well over a year as well as the centrally determined HER2 scoring of archival tumor samples. In conclusion, on Slide #8, preliminary data suggests that evorpacept can be safely combined with no maximum tolerated dose reached in combination with trastuzumab, ramucirumab and paclitaxel. In combination, evorpacept demonstrates a durable 72.2% objective response rate in HER2-positive second-line gastric patients with a 17.1 month median OS that suggests clinical activity beyond that expected from relevant historic controls. I'm now honored to introduce to you on Slide #9. Dr. Kevin Harrington, who will speak about evorpacept for patients with advanced checkpoint naive head and neck squamous cell carcinoma. Dr. Harrington is a Professor of Biological Cancer Therapies and the Head of the Division of Radio Therapy and Imaging at the Institute of Cancer Research in London. He leads the targeted physical therapy section within the Royal Marsden ICR biomedical research center and is the National Chair of the CRUK ART-NET Group. We are the beneficiaries of his deep experience in biologically targeted approaches to cancer therapy and especially head and neck squamous cell carcinoma. Kevin?

Thank you, Sophia. Today, I'm really pleased to be able to speak on evorpacept for patients with advanced head and neck squamous cell carcinoma. On Slide 10, you can see that head and neck squamous cell carcinoma is also 1 of the common global cancers. There is worsening prognosis once in the advanced setting. Its incident rate is highest in Eastern Asia, but there are increasing frequencies of this disease in the United States and in North America. On Slide 11, I show you that pembrolizumab was initially approved as a monotherapy in the second-line setting based upon the pivotal KEYNOTE-040 study. It was subsequently studied in the advanced first-line head and neck cancer setting in the KEYNOTE-048 pivotal study, and it is now considered standard of care both in the first line setting where it can be used both as a monotherapy in patients with PD-L1 positive disease and in combination with 5-FU and platinum in patients with head and neck squamous cell carcinoma irrespective of their PD-L1 status. One challenge that we encounter with immuno-oncology agents in this disease, and this is demonstrated in the table has been the identification of Phase II endpoints that are predictive of overall survival. That, of course, is the global standard of clinical benefit for patients with head and neck squamous cell carcinoma. As evidenced in both the KEYNOTE-040 and the KEYNOTE-048 studies, short-term response metrics such as initial overall response rate or even median progression-free survival have not always predicted the eventual clinical benefit as measured by the median overall survival seen with pembrolizumab monotherapy or in combination with the fluoropyrimidine and platinum in the first- and second-line settings. Indeed, in Slide 12, the separation of the experimental arm for the control arm in the Kaplan-Meier plots of the total trial population for KEYNOTE-048 and for 040 for both progression-free and overall survival does reflect the ultimate benefit of pembrolizumab containing combination compared to the control arm. But as you can see on Slide 12, the benefit improves over time when looking at either progression-free or overall survival. However, in the case of the median progression-free survival in both studies, the improvement in clinical benefit is not really -- is not yet seen between the test and the control arms at the median point. It is not until more progression-free survival or overall survival events have occurred reflected at later time points, including the median overall survival, the overall survival rate at 12-month landmark analysis or the OS rate at the 24-month landmark analysis, this is where the clinical benefit is increasingly demonstrated. I am now pleased to present the updated results from ASPEN-01 on behalf of the study investigators. I will present Phase I data as of September 1, 2021, summarizing evorpacept for patients with advanced head and neck squamous cell carcinoma presented at the annual meeting of SITC 2021. On Slide 14, I show you that evorpacept was administered with standard pembrolizumab, 5FU and platinum in patients with first-line head and neck squamous cell carcinoma, or with pembrolizumab alone in patients with checkpoint naive second line or greater head and neck squamous cell carcinoma. The primary study endpoint of this portion of the study was determination of the maximum tolerated dose of evorpacept in combination. The study enrolled patients from the United States and South Korea, with more Asian subjects enrolled into this cohort. All patients enrolled into these 2 cohorts have had no prior treatment with a checkpoint inhibitor for their head and neck squamous cell carcinoma. As shown on Slide 15, the most common evorpacept-related adverse events in combination with pembrolizumab and chemotherapy, occurred in 1 patient each and these were neutropenia, fatigue, pneumonitis and anemia. The adverse event profile of patients receiving evorpacept and pembrolizumab has been reported elsewhere. There were no dose-limiting toxicities, no on-study deaths nor evorpacept-related serious adverse events in this cohort. The maximum administered dose in combination was 15 milligrams per kilogram every week. On Slide 16, you can see that in patients with first-line head and neck squamous cell carcinoma, receiving evorpacept in combination with pembrolizumab chemotherapy, a 38.5% overall response rate was observed in patients with both high and low PD-L1 expression. A median progression-free survival of 5.6 months, a median overall survival that was not reached and 85 -- 87.5% overall survival rate at 12 months compare favorably to the pembrolizumab chemotherapy arm of KEYNOTE-048 as a historical experience. In that situation, there was a 4.9-month median progression-free survival, a median overall survival of 13 months and 53% estimated overall survival rate at the 12-month landmark point. Additional follow-up of enrolled patients with second-line checkpoint naive head and neck squamous cell carcinoma who received evorpacept and pembrolizumab showed that there was an overall response rate of 40%. Here, there is a 4.6 month median progression-free survival, a 24.5 month median overall survival and 80% estimated overall survival rate at 12 months. And these compare favorably to the pembrolizumab KEYNOTE-040 historical experience. In that situation, there was a 2.1-month median progression-free survival, a median overall survival of 8.7 months and an estimated 40% overall survival rate at 12 months. Again, these data were reported in similar populations. So in conclusion, on Slide 17, a I've shown you that preliminary data suggests that evorpacept can be safely combined with no maximum tolerated dose reached in combination with the regimens tested. Patients with checkpoint-naive first line and second line or greater head and neck squamous cell carcinoma who were administered evorpacept combination reported clinical benefit beyond historical landmarks, especially when looking at survival-based end points of median overall survival and overall survival at 12 months as measures of clinical benefit. I thank you for your attention, and I will now hand the presentation back to Jeanne. Jeanne?

Thank you, Dr. Harrington. That concludes our prepared remarks. We'd like to open the line for questions. Operator?

[Operator Instructions] Our first question is from Michael Yee with Jefferies.

Thank you for the summary of the data. We had 2 questions. First on head and neck data, appreciating. I think you sort of explained that you can't always look at overall response rate. And here, there's often a tail for improved survival in immuno-oncology mechanisms. Can you just still attempt to put into context your 39% overall response rate versus the others and relate that to your mechanism of action, which is a don't eat me mechanism in which you would still think there's more cell killing. Can you maybe just put into context what you think is going on there, given the mechanism and also in the context of small numbers for survival? That's my first question. The second question, if I may, I just have to sneak in an MDS question. So if you could comment on the MDS abstract that would be helpful as well as obviously investors are trying to put that in the context as well. So please.

Yes. This is Sophia Randolph from ALX. I think as -- first off, and then I'll ask Dr. Harrington to also speak again to the concept of response rates and how we think about that in this particular disease setting. But first off, in terms of the data from our head and neck cohorts. We are actually very encouraged with this data. Again, as you look at how immuno-oncology agents in general seem to be working in the head and neck space. It does seem as though the impact of immuno-oncology agents does take time for those clinical benefits to be seen. So while in the short term, very similar to the pembrolizumab experience in KEYNOTE, we do see objective response rates improvements. Although, again, we see what we see, the 38.5% in the first line and the 40% in the second line. Certainly, that -- those responses do appear to deepen over time. And that impact of that durability and the impact of other patients even with stable diseases over time seems to be what's driving the longer PFS and particularly overall survival. The drug itself as it works in combination with T cell checkpoint inhibitors. It's really enhancing dendritic cell cross-priming of T cells. So that in it of itself leads to an increased activation of cytotoxic T cells. And then when those are in the presence of its T cell checkpoint inhibitor, namely pembrolizumab, we can see this increased activity. So it may be a little bit different than the kinetics of what we see when we're combining with an anticancer targeted agent. But really, as we get into larger trials, and this speaks to the size that you had asked earlier, as we get into the larger randomized Phase II setting. We hope that we'll be able to learn more about how this drug works in that setting. I'm going to turn it over to Dr. Harrington, if you want to speak a little bit more to this concept of objective response rates in these diseases.

Yes. Thanks very much, indeed, Sophia. I think you've given a really nice summary of the data. I guess, having been an investigator both in the KEYNOTE-040 and in the KEYNOTE-048 studies, we do recognize the fact that there are patients in whom you don't see a response, but in whom you subsequently see long-term survival. That is undoubtedly true, but also, we see that patients who derive a response, especially I remind you of the KEYNOTE-048 data with single-agent pembrolizumab, the median duration of response of responding patients was nearly 2 years. So responding per se is important because it is associated potentially with a durable remission and therefore, that can read out into durable and long-term overall survival. So taking that context, I think the data that we see here and I'll do it in reverse order in the second line setting with an overall response rate of 40%. Well, that's -- albeit with small numbers. You asked me to comment on the small numbers, we can't get away from that. Nonetheless, as you know, the best response that we saw in single agent pembrolizumab in KEYNOTE-048 was 23% for the CPS greater than or equal to 20%. So if this were replicated in the randomized trial, this would be far and away above what was witnessed in the KEYNOTE-048 study. And in the first-line setting, of course, those response rates look a little bit closer to what we saw, for instance, in the patients receiving chemotherapy plus pembrolizumab in KEYNOTE-048. But again, I think the survival rate at 12 months at 5 out of 6 patients, 87% is really very compelling and it's very interesting. And it's certainly what's driving me to be interested in being an investigator in this trial. I think the mechanism of action has been covered by Sophia. And I think the combination with chemotherapy, of course, gives us that additional benefit of potentially triggering immunogenic cell death at the same time as having an anti don't-eat-me signal on board or a blocker of the don't-eat-me signal, which, of course, can promote greater degrees of ADCP and then, of course, that cross-priming event which leads to T cell activation, which in the presence of the anti-PD1 antibody hopefully will translate into better responses and subsequently into more durable responses and subsequently to improved overall survival. So I think those are the really key points that I would take from these data while recognizing n is small in each of the components of this analysis.

That's very helpful. And Sophia, you're going to get the question anyway. So that question around interpreting the MDS abstract which the markets seem to digest in a negative way. Can you just put some color around that how to compare that to [indiscernible]?

Sure. Yes. So just very briefly, the ALX148 for its updated data in looking at evorpacept plus azacitidine in patients with first-line MDS and relapsed/refractory MDS was reported out in abstract form a couple of days ago. And I think in that case, we also put out a press release, and I think there was a little bit of a mix up between folks who were looking at the abstract and not taking into account the press release. In the abstract, we reported 2 out of 5 patients with first-line treatment naive MDS who had, had cytogenetic response As we got the complete data set on those patients after the abstract was submitted, those 2 patients also met criteria for CRs. And so that updated data was presented in the press release that went out with the accompanying abstract. So overall, in the 5 evaluable patients who received first -- in the first line treatment untreated MDS population, we had 2 out of 5 patients with CRs, which is -- it's small numbers, but if you want to put the percentage to it, that's 40%, which is in our minds, certainly demonstrating clinical activity and on par with some of [magrolimab] earlier data, certainly, an improvement upon the backbone of azacitidine where in that population, you could anticipate about somewhere around 17% to 20% objective response rate. These patients, just by chance were all TP53 positive, TP53 mutation positive, the -- including the 2 responders. The trial does enroll in all comers population, and we'll be presenting additional data at ASH if -- in approximately 20 patients overall. On the relapsed/refractory side, we also had 5 evaluable patients who are described in that abstract. In this population, one might expect to see really anything in the relapsed/refractory setting is completely significant. 2 out of the 5 patients had no CRs. So again, very encouraging and we'll be presenting the full data set at ASH in the another few week.

Our next question comes from Chris Raymond with Piper Sandler.

This is Nicole Gabreski on for Chris. So maybe just kind of touching on ASPEN-03 and ASPEN-04. It looks like for your updated corporate deck that you've shifted the endpoints to now include a 12-month OS rate as well as response rate. So I guess maybe we just had a couple of questions based on this. First, how should we be thinking about what success looks like given the co-primary end points? And on the response rate portion, do those rates need to be in line with other comps? Or what kind of benefit will you need to show in your Phase II studies to be meaningful? And then secondly, can you just kind of talk about any regulatory interactions you've had in regards to modifying the primary endpoint and what the feedback has been?

Got it. Before I get to the details of the stat design, I'm going to ask Dr. Harrington if you want to speak to just the concept of why we incorporated the 12-month OS rate and -- into those studies.

Yes. I mean I think as a key secondary endpoint in each of those studies, I think overall survival of course, is what people are looking for. And clearly, what the -- those that register drugs, the FDA and outside of America, the EMA are looking at. And so I think it's important that we get a sense of how active these regimens are and look in terms of overall survival. I think we get a very clear indication. If you look at the data from CheckMate 141, you look at CheckMate -- KEYNOTE-040 and then a KEYNOTE-048 in the first-line setting, we see that even in situations where curves cross one another for OS. So we do see that in the first-line setting in 048. We see that early on, and especially in those tumors that are PD-L1 on the low side, we see crossing of the curves. Those curves have crossed really no later than 7 months and that we see that the shape and the direction of travel of the overall survival curves that establishes itself at about 12 months is then feeding into what we see as the longer-term tail on the curve as it emerges. So I think the earliest that you get a clear indication of the direction of travel and what the likely benefit and the separation between curves is going to be is around that 12-month point. To try to look any earlier, I think we'll give you potentially a false view of how things are going to emerge. Clearly, the longer you leave it, the clearer the picture becomes and that's the reason why we publish long-term data from these trials. But I think when you're looking for a signal with a view to deciding on powering a Phase III study or making strategic decisions about how to develop an asset, I think overall survival of 12 months really is a very key and powerful endpoint. I hope that addresses the question that you had.

And thanks, Kevin, and this is Sophia. And I can also speak a little bit to the statistical design. So with what Kevin has said. How we've incorporated this into both ASPEN-03 and into ASPEN-04 is to create a statistical design where we have co-primary endpoints. So we start by looking at that 12-month OS rate first. And so the trial has increased a little bit in size due to the powering around now incorporating both a 12-month OS rate and the co-primary objective response rate. So now the size of our trial is around 177 patients, and we'll be evaluating these different co-primaries in a fixed sequential design. The overall structure of the trial is the exact same as what it was before. It's still a 2:1 randomization. ALX has been given us 45 mgs per kg to 3 weeks to make it more convenient to be combining and because of our safety profile, we can do this to be combined with standard of care keytruda in terms of 03 and in terms of 04 in combination with KEYTRUDA, either on cisplatin. So in terms of what does success look like for that 12-month OS rate in the KEYNOTE-048 first line head and neck 12-month OS rate was around 51%. So we'd be looking for an improvement there to somewhere in the mid-60s is what we've powered the study to look for. And then in terms of the OR rate as a fixed sequential design, if that OS rate is positive when we go on to look at the OR rate? But certainly, for the OS rate, that's going to be really the key of those 2 endpoints, the first one that will be evaluated. And then in terms of the ALX keytruda chemo combination in ASPEN-04, there in the KEYNOTE study 12 OS rate, backbone is around 53%. So fairly similar. And again, looking to improve upon that into the around 66% or the higher 60s as a surrogate of what we believe it will be a surrogate of clinical benefit as measured by overall survival, which would be a key secondary endpoint as Kevin mentioned. Hopefully, that addresses your question. And some additional detail around that will be provided in the trials and progress case studies that we have for both ASPEN-03 and ASPEN-04 that will be presented at SITC as well.

Our next question comes from Brad Canino with Stifel.

Thanks for hosting this presentation. For Dr. Harrington, it looks like only 2 patients progressed rapidly in the frontline head and neck quad. Can you comment at all about how that compares to the keytruda chemo combo, would you expect more rapid progressors without evorpacept? And then I have a follow-up.

Thank you for the question. So the marked difference in the KEYNOTE-048 data sets between the single-agent pembro and the pembro chemo combo is really the shape of that early part of the curve as you correctly I guess, suggesting your question. The reality is that the -- for the pembro chemo combination, there isn't a clear detriment early on from OS, which is what we see with single agent pembro. So what would I expect, I guess, is difficult to know. But we would expect relatively small numbers of patients if they are covered with the combination of chemotherapy. And that, to me, is 1 of the main attractions of using chemotherapy pembrolizumab combination regimens and especially in patients with bulky disease with symptomatic disease where even relatively small volume progression before you evolve a response can be associated with a significant detriment in terms of quality of life or indeed even pose risks to the patient in terms of getting pneumonia or something like that from swallowing dysfunction that could lead to an early mortality event. So I don't think there's enough data for us to make a clear comparison here. What we can say is that evorpacept is certainly not performing any worse than this. And one would hope that it would indeed guide against any of those early progressions and hopefully would mitigate the risk of that. I don't think we have enough data to say conclusively that that's happening.

Great. And then for Jaume or Sophia, a strategy question for you because now that you're moving away from the ORR endpoint from the ongoing frontline randomized Phase II trial are you more likely to now need a Phase III trial for approval? Because I can't really think of any precedents for surrogate approvals on a 12-month OS. And then I guess, can you talk through any updates to the data readout timing, which I think you previously suggested might have been in 2023?

Yes. This is Sophia, I'll start, and then Jaume can speak together on the milestones. I think in terms of the strategy for our or combinations in head and neck. Yes, I mean, as we have discussed before, the original design of this study, was potentially registrational with the OR rate. We certainly haven't taken the OR rate out of the design. We've only added in an additional co-primary endpoint. So I think in terms of the design of the study, ultimately, it will be a review issue as it always is with the agency in terms of looking at the data and they'll have to take into account the landmark analysis of OS and to your point, in this field, median overall survival has always been the benchmark for full approval. So it will have to be an evaluation of the 12-month OS rate. However, the original OR rate is still there. And as we have discussed before, depending on the data on that trial design had on -- had still -- would still considered as potentially registrational from the agency standpoint. I think regardless, we're going to have to do a Phase III, whether regardless of whether or not something meets criteria for accelerated approval or not. So we certainly are moving forward with that in mind. And are working on Phase III development plans as we would anyway, in the development of this agent in head and neck. Hopefully, that addresses your question.

Yes, in terms of complete[indiscernible] status.

Yes. So for the milestones for this, they're all slightly larger studies. So I think we've gone from size of 105 with the ASPEN-03 study to 177. And then from the ASPEN-04 KEYNOTE combo study going from 106 up to 162. The actual overall time lines aren't any different because we can operationally address that with just additional sites. So the -- there is an added -- just with the 12-month OS rate, there is an added follow-up period that has to happen. However, with an operational approach, we don't anticipate the overall time lines for these Phase IIs to change.

Okay. So it sounds like you get accrual or at least are targeting get accrual by the end of next year to kind of get the 12-month follow-up and then have data by the end of 2023. Is that fair?

I think what we've said externally is just when we anticipate getting a readout from the study and those time lines have not shifted.

Yes. The only thing I would add is that we should actually look at these 2 studies in a different way. I think the strong debulking effect of chemotherapy is masking a little bit or effect, or are effect or pembro effect in the case that we do not have chemotherapy in the case of the pembro combination only, I think that will be easier to see OR as we're seeing now, the smaller of patients we have.

And our next question comes from Colin Bristow with UBS.

So in the head and neck data, is there any meaningful difference in the baseline characteristics or dose exposure in the additional 9 patients that you would highlight ultimately grow the 22% OR versus the 75% in the first 4. And then secondly, just with regards to OS rate, is it reasonable pining on this and making changes to the Phase II trials at this stage, just given the maturity of the data, I think the beating follow-up is 6.2 months? And then just one final question. Have you thought about potentially enriching for even higher CPS scores in 3 and 4, given it looks like there is a correlation here? Or is there a concern that you'd then be operating at in end of the wedge in terms of the incremental efficacy you could demonstrate on top of Keytruda?

So I can start with talking a little bit about the ASPEN-01 data itself. So as you asked about the additional patients who have been enrolled, honestly, the overall patient population it haven't changed that, like we still have a look through here. We still have about in both cases, about somewhere around 54% to 60% of patients who have this role, distant metastases who -- which is a marked aggressiveness of the disease, the relative enrollment of now versus soon now the predominance of Asian enrollment from our South Korea colleagues. Those are been fairly similar to what we had presented in the second line setting, what we're seeing on the first line setting. So I don't think there's anything to be renewed there. In terms of CPS scores for the client data set, it's a very small data set, so it isn't all comers. And which is consistent with pembrolizumab label in combination of chemotherapy, it is based on a CPS agnostic population. And so here, when you look at the total population, we actually are seeing a nice mix of both CPS low as well as CPS high in the ASPEN-01 data set. And are they with small numbers, we have seen a complete response with the patient with -- on the waterfall parts you can see if you go back and look with CPS score of 50 and a very deep response in the patient with the CPS score zero. So both at high and low CPS scores, we're seeing responses, but again, small numbers. So as we go into the ASPEN-03 and 04 larger randomized Phase IIs, CPS is certainly important. And as part of the randomization, we'll be taking into account certain sort of stratification like factors, one of those will be CPS score. And so patients with different CPS as they fall into the different CPS buckets had a slightly different characteristics in terms of their responses in the KEYNOTE-048. So CPS score is important, and we'll be stratifying for that to be stratifying by region, so Asia versus ex-Asia as well as looking at a few other stratification factors as well. I'm going to hold off there and Dr. Harrington, I don't know if you want to speak to just in general, the concept of some of the characteristics of a population that can have an impact upon response that was from the KEYNOTE-048 series.

Yes. I mean I think it's a very interesting conceptual question about would we a priority expect that patients who are CPS high or those who are CPS low might derive the greater benefit from the addition of evorpacept to pembrolizumab with or without chemotherapy. And the honest truth is I think you could probably construct a line of sight theoretically to either of those groups being likely to derive the greater degree of benefit. Honest truth is we don't yet know. I think I'm attracted to the idea that the addition of chemotherapy with triggering cell death and enhancing phagocytosis and in post presentation is a good way of going I think also for those patients in whom you have the absence of chemotherapy and whereby you lying upon or percept to add something else to pembrolizumab, where pembrolizumab itself may be insufficient, again, is an attractive hypothesis. And I think we'll see this from the data, and I think the trial is designed in such a way that we'll get a good indication from that. In terms of the distribution of the patients that were in the original ASPEN-01 study, I think it's difficult to draw any conclusions. I think certainly, this isn't what we're used to seeing in registrational studies, certainly in terms of the distribution and the predominance of patients coming from Asian jurisdictions. That's not something we normally see in registration studies, at least not those that we take part with our European and North American colleagues. But I don't think necessarily we can conclude that the data are invalid for that reason, and I certainly would not come to that conclusion. More generally, in terms of the distribution of patients in KEYNOTE-048 and how that can impact. And of course, irrespective of PD-L1 CPS value, and that does have an impact on response rate and overall survival. I think the other things that we do need to be aware of, of course, are the patient performance status, which is really very important and obviously also P16 status, HBV status of the patient. And in terms of the performance status, I mean, that gives us a clues to the disease burden and the likelihood of symptoms arising from that disease. And we know that patients with bulkier disease, most likely and most investigators would probably agree with this will drive greater degrees of benefit from richens containing chemotherapy because it gives them that debulking of disease that you don't necessarily get early on with single-agent immunotherapy. I guess those are the only additional comments I'd make Sophia.

Yes. Thanks. And those other factors that you mentioned, P16 and performance status are also part of the stratification design for asking that you were asking for.

And just on the idea of pinning on and making changes just based on relatively immature OS rate data.

I'm sorry, say that question one more time.

I just -- what we've been talking about OS rate and making changes to Phase II trials on the back of a 6.2 median follow-up in the head and neck cohort. I'm just curious how do you guys feel about that?

I see. So no change has been made to the ASPEN-01 study, but based where we're seeing that follow-up. But based upon that -- the information that is coming in appropriate or so, based upon the information coming out of this first in human, we have modified the 2 randomized Phase II studies that have just begun. For each of those randomized Phase II studies, there is a safety lead-in portion, which is the initial portion of those studies, which is a nonrandomized safety lead-in confirmation cohort because we're looking at the 45 mg per kg [Q] 3 week dosing of ALX for the first time in these studies. So the changes to the randomized portion of the study, we have not entered into the randomized portion of the study yet. So this is -- that's an important point, which I'm glad you raised that. So we are not changing the design of the study after having started the randomized portion. This is a prospective change that will go into effect after we complete the safety lead-in portion of the study, just to make that clear.

And our next question comes from Alethia Young with Cantor Fitzgerald.

A lot have been asked. I guess maybe one is just talking a little bit about kind of how the data kind of continue to hold up, both in gastric and head and neck. How do you think it consolidates kind of second line and beyond since they are like kind of many treatments used there? And then also just looking at the safety. It continues to look obviously differentiates. So I just wanted to kind of just get any perspective that you have there? And how does that -- does it further kind of increase your conviction across other solid tumors?

Thanks, Alethia. So for the gastric cancer, I mean, we haven't had a chance to speak much about the gastric cancer. The gastric cancer data coming out of the ASPEN-01 continues to improve and continues to remain strong. The last time we presented this data was at ESMO GI, not too long ago, about 6 months ago. And so the update that we're providing at SITC, we see even still further improvements in the overall response rate, but particularly a 72.2%. But particularly in the duration of response, where we're seeing our 14.8%. And then starting to see that now impact dragging out the median PFSs and median overall survival end point is, of course, the gold standard in the aggressive disease. So in terms of the question around how does this fit with the other second-line agents that are being used in the field? I think comparatively very well, ram-pac is still the global standard in the U.S., trastuzumab, deruxtecan is being used in the second-line setting based off of the third-line data. The preliminary data that we're showing here really shows a very nice improvement over both of those backbones of therapy. And so for us, we're very -- again, very excited to move this forward in the second-line setting. There has been a smaller study, as I mentioned before, a single arm, predominantly single institution looking at tras/ram/pac. And that in and of itself with an objective response rate of 52% and median PFS of 7.4. We're seeing an improvement over that with the contribution of evorpacept on top of that triplet backbone. Again, as much as you can compare between 2 different studies and single arms and single institutions. But I think that the buffer that we're seeing in all of these end points compared to the current second-line therapies that are in play right now. The quadruple of evorpacept plus tras/ram/pac, preliminary data suggests that we're seeing superior activity. The proof of that, of course, won't come until we go into a randomized setting, and that is our next step. So to directly address the contribution of evorpacept to tras/ram/pac factor, the next step in ASPEN-06 will be a Phase II study randomized to look at the quadruput, evorpacept tras/ram/pac versus tras/ram/pac and that will be the next step. And then after that, we'll do the Phase III, which will be comparing the evorpacept, tras/ram/pac to the regulatory comparator, which would be [indiscernible]. I think I answered your question. The deep -- I think that was the other. Yes. For safety profile across -- honestly, across the program, it continues to be rock solid. I mean, we've treated now cross program over 185 patients. We're starting to get into new combinations in the myeloid space, combining with azacitidine and we've just initiated dosing in our AML study looking at evorpacept [indiscernible]. So whether there or whether you're on the solid tumor side here with multi-agent chemotherapy regimens, checkpoint, antibody, the safety profile has been -- continues to be unremarkable. And that's, I think, one clear differentiator for evorpacept versus other CD47-targeted agents. And it does allow us to be able to combine with these chemotherapy-based or other cytotoxic based therapies and move this drug up earlier in line of therapy where we can potentially have the greatest benefit upon patient's course of their disease and the safety profile allows us to do that in a way that I think other CD47 targeted agents will be challenged. So that's definitely from a development standpoint, that safety profile is something that we believe makes us to the best in class molecule.

Maybe one follow-up. And I get the question every so often, even sometimes from physicians, just that maybe the safety means more in solid tumor versus in some of the hematologic malignancies like MDS, for example. I mean, do you guys believe that to be true? Or like how would you -- what's your perspective on that?

Yes, that is interesting. Even in the myeloid space, people die from complications of cytopenia. So I guess I would argue, if you have a drug that doesn't exacerbate the cytopenia associated not only with the disease itself, but also with current treatments out there, backbone of [indiscernible], that would be a good thing. And on the solid tumor side, I think that benefit is very clear. On the heme side, again, anything that can decrease potential complications due to cytopenia can decrease blockage or risk for to have multiple fusions. I think that, that's a good thing and something that the patients could benefit from. We just don't want it to come at the expense of efficacy. And certainly, to date, we are not seeing that with a very preliminary data.

And our next question comes from Adam Evertts with LifeSci Capital.

Thanks for the data update, gastric data, in particular, quite impressive. I just want to go back to the combination of evorpacept and anti-PD-1. So on the topic of dendritic cell priming and T cell activation, do you have or will you have any translational data to support or outline that piece of the mechanism in head and neck?

So in terms of the -- I'll speak to what we're doing in the clinical space. And then Jaume, maybe you can speak to just what we've learned through our nonclinical evaluations with checkpoint inhibitors. In the clinical space, the pharmacodynamic endpoints that we have been focusing on -- have been looking at receptor CD47 receptor occupancy. So we've looking at the ability of the drug to access its target not only in the peripheral blood, but ultimately into the bone marrow itself and that data that we we'll be evaluating in the Phase II study. Where we'll have a much more homogeneous population of patients to evaluate. And then doing some characterization of the immune cells that we see in the different compartments. So different ways of looking at T cell activation and tumor infiltrating lymphocytes or tumor infiltrating massive arteries into these different compartments. So in the clinical setting, those are some of the things that we'll be looking at. In terms of the general mechanism of action, Jaume, do you want to speak to that?

Yes. So from the clinical setting, and we have published -- we have seen that the blockade of CD47 T cells [indiscernible] and therefore, they are constantly inhibited by interaction with CD47 on cancer cells, we have seen that evorpacept alone by itself thus activated the T cells, and these [indiscernible] cells can activate [indiscernible]. And these T cells obviously in combination with our PD-1 can be even more active against tumors. Another mechanism that we have shown the pre-clinical settings, we and others have shown in pre-clinical setting for CD47 blockade is the conversion of M2 macrophages into M1 macrophages, therefore, making them less immunosuppressive and more anticancer in the tumor macro environment, and this data has been published and can be found in our website.

And I'm not showing any further questions in the queue. You may continue with any final remarks.

If there are no more questions, we'd like to extend our appreciation to Dr. Harrington for his valuable insights today, and thank you for your participation. Have a good day. Thanks.

And with that, we end our program for today. Thank you for your participation. Have a great day.