All right. Great. Good afternoon, everyone. I hope you had a great weekend. Tyler Van Buren here, Senior Biotech Analyst at TD Cowen. Thank you very much for attending TD Cowen's 6th Annual Oncology Innovation Summit. For our next session, very excited to have a fireside chat with Arcellx. And it's my pleasure to introduce Rami Elghandour, the Chairman and Chief Executive Officer, Chris Heery, the Chief Medical Officer, and Michelle Gilson, the Chief Financial Officer. Rami, Chris, Michelle, it's a privilege to have you here. Thank you for joining me.
Thanks, Tyler. We're excited to be here.
Before we get started, for those in the audience, you can feel free to email me questions at Tyler.VanBuren@tdsecurities.com or Tyler.VanBuren@cowen.com. I'll do my best to get them asked before the end of our discussion. With that, we'll go ahead and get right into it. You'll surprise us with a press release of the data that you are close enough to the data that you plan to present at EHA. Maybe you could start by recapping some of the key highlights from that top-line data release that we saw just the other day.
Yeah, sure. Thanks, Tyler. Maybe I'll kick off here and then pass it over to Chris. Look, we feel like this data set from our iMMagine-1 registrational trial really defines the profile of anito-cel as the therapy of choice for patients in the myeloma community. It is really based on that unique combination of attributes across safety, efficacy, and scale, and reliable manufacturability. The reason we released it surprised you, as you said, is because we do believe this is a really impactful data set. We wanted to make sure we had the opportunity to discuss that with the investment community. With that background, I'll turn it over to Chris, and he can talk through the highlights.
Yeah, thanks, Rami. I think the good news is that we continue to look like the data are trending exactly the way we expected them to when we presented the first snapshot at ASH. The complete response rate continues to mature with all of the patients now included in the evaluable group for efficacy with a 68% complete response rate. With a median follow-up of about 12 and a half months, we think that is directly in line with what's been seen with Carvykti and puts us on an efficacy front in the best-in-class space with that amount of follow-up for that complete response rate. That's supported by the PFS rates at six and 12 months, which are also directly in line with what's been observed in the Kaplan-Meier estimates at six and 12 months for Carvykti.
Maybe more importantly, with that efficacy, we have still not seen a case of a delayed neurotoxicity event, including Parkinsonism, cranial nerve palsies. We haven't seen any enterocolitis. We do think that that is a distinguishing factor between the products at this point. We've heard it said that this is a class effect, but we just don't believe the data support that observation or that conclusion. When you add those two things, the safety and efficacy profile of anito-cel to the fact that Kite will be manufacturing this product and already is in iMMagine-3, we believe that this will be the best product available for patients seeking a CAR-T with relapsed refractory multiple myeloma.
Great. Thanks very much for that overview. With the full presentation at EHA in a couple of weeks, should we expect to receive preliminary PFS curves? Will it be the same data cutoff? What else should we expect from the presentation?
Do you want me to take that, Rami?
Yeah.
Yeah. In terms of PFS curves, it's been our policy, including in our phase one, to not use Kaplan-Meier curves too early because the tail of those curves tends to be very unreliable. What we think the challenge with looking at Kaplan-Meier curves right now is that if that tail is unreliable, people will automatically want to compare it to CARTITUDE-1. That's really impossible at this point because there are not enough patients with enough follow-up beyond 18 months, for instance, for it to be a reliable curve out at that point. That is why we do the snapshots of Kaplan-Meier estimates at six and 12 months because they give a good indication of the shape of that curve where we believe it is already reliable. Looking at, for instance, things like subgroup analyses have the same weaknesses at this early follow-up.
There's just not enough data and enough follow-up for those patients to be able to say reliably. We know that while there will be more opportunities to update the data, we think the level of interest in these data is such that people may want to latch on to things a little too early. At this point, we will most likely be sticking with what was disclosed at ASH, just updated with more patients, more data. You'll get some of the same information that was in the press release. There will obviously be a bit more nuance, things like demographics. You'll get those data. We'll have more data listed in terms of adverse events, et cetera. We included in the top line in the press release all the things we thought were most critical.
Maybe I'll just add one other thing, Tyler. I think as we've spoken before here on the release of this data, we're comparing that in terms of follow-up to the ASH 2020 presentation from Carvykti. I think if you look at that data set, the profile of Carvykti hasn't really changed since that presentation with that much follow-up in terms of its efficacy profile, which obviously remains very strong. Some of the safety concerns that were present at that time, and certainly the manufacturing turnaround times, are largely consistent with that time period. That's part of the reason we felt like this was a fairly defining data set because based on precedent, at this degree of follow-up, this data tends to define the overall profile of a drug on a go-forward basis.
I guess based upon that commentary, I'll skip forward to just a question again on safety. We keep on asking this question and thinking about whether we're going to see cases of delayed neurotox or Parkinsonism, but we keep on seeing no cases. Given the follow-up that we now have, what do you believe is the probability that we will end up seeing a case? Even if we do see a case, how do you think it will compare in terms of the rates that we've seen with Carvykti, right, given the denominator in terms of the number of patients that have been treated?
Yeah, I mean, look, I think it's very clear at this point that these are very different therapies from a safety profile perspective. While this is not a perfect analogy, this is like continuing to ask an electric car manufacturer if they're going to have oil leaks in the future. I mean, these are fundamentally different technologies and different platforms. As a result, the profiles are different. Now, I can't—that's why I said it's not a perfect analogy because we have seen, even with Abecma, there has been a splattering of cases here or there, just given how sometimes these things can be attributed. I think at the end of the day, it's very clear that these therapies are different from a safety profile, at least on the evidence we have to date.
Even if we were to have one or three or whatever number of cases in the future, we do not believe that would change that profile. I do not know if Chris, you want to add anything else to that.
I think you covered it really well. I think when we look at some of the data that are now becoming available at ASCO in abstract form and some of those that have been available over the last six to 12 months in the real world, we see a pretty consistent pattern despite modifications of management, despite all sorts of attempts at intervention. The rate of Parkinsonism with Carvykti seems to be right around 4%. The rate of these gastro—sorry, these colitis and rhinitis events are somewhere in the range of 4%, and the cranial nerve palsies are somewhere in the range of about 5%. Those things, at the current number of patients that we have dosed with the amount of follow-up, would become nearly statistically impossible to have not observed by now. Not quite impossible, but very, very unlikely.
Again, to answer your question, it doesn't mean that we'll never have some challenge where a patient has one of the symptoms of one of these things or has something that looks like maybe it's near it, right? When we look at how it's been described with Carvykti at these rates and the onset of symptoms and how progressively damaging they can be to the patients, those aren't things that should be easy to miss, and they should have been observed by now if they existed.
Okay. That's clear. As we think about ALC or absolute lymphocyte count, are you guys monitoring for ALC, or have you monitored for ALC in any of the cell trials?
No. Never.
We capture it, of course, so we can look at it from an analysis standpoint. We do not make any intervention based on ALC. We look at, does a patient have CRS? If they do, are there things like inflammatory markers and things like that that are associated with those? These are all parts of guidance that are generalized documents that are available to anybody and are recommendations from multiple third-party groups that have recommended how to manage these cases. We follow those guidance documents.
Yeah. Just to be clear, we only really looked at ALC in this context after it was disclosed that Carvykti was using it as sort of a surrogate for these delayed neurotoxicities. Naturally, we went back and looked at our data and said, "Hey, do we see a secondary rise in ALC or any of sort of the characteristic antigen-independent expansion?" We did not see that from a PK perspective, but it was not something that we looked at during the conduct of iMMagine-1 or that we are using in the execution of iMMagine-3.
Okay. Have you guys excluded patients with peripheral neuropathy from Velcade or for other reasons?
No.
That would be almost impossible to do if you're going to enroll a trial in relapsed refractory multiple myeloma.
Okay.
I'm guessing that somewhere in the range of 30%-50% of patients after Velcade are going to have some level of peripheral neuropathy, usually grade 1. It's usually relatively mild, but almost everyone's going to have something at some point. It would be nearly impossible to enroll a study in this space if you were going to prevent enrollment of that group of patients.
Okay.
The answer is no.
All right. Just to be clear, use of steroids or tocilizumab in the trials has been pretty consistent historically in what you're using in iMMagine-3 and moving forward.
Yeah. What we presented at ASH, which is that we do believe that when a patient has a fever within 48 hours of dose, that those patients do pose a higher risk because most of those patients have higher disease burden, that those patients should get Toci and Dex immediately if they have a fever in the first 48 hours. That's obviously not the majority of patients. Most of our patients did still get Dex with their first dose of Toci, just as sort of a standard approach. The large majority of our patients only ever received a single dose of dexamethasone. I think what that shows you is that it's not really the steroids that are the differentiator here. I think on the edge cases, they can prevent, say, a grade 1 CRS maybe from being a grade 2.
But that's really the edge cases, right? So of all patients that have grade 1 CRS, maybe only 5%-10% of them might have drifted to grade 2 if you didn't intervene like that. But I think more importantly, the idea that giving those doses of Toci and Dex could have prevented things like delayed neurotoxicity is really not supported by anything we've seen in the literature or by what we've had discussions with multiple clinicians about their experience. So I think the other way to think about this is if that could prevent delayed neurotoxicity, then it seems like a very simple solution one could use to prevent it with Carvykti. But we also know that this is the approach that most of our clinicians and the biggest centers, most of the clinicians that enrolled our study at the biggest centers are already doing for Carvykti.
I don't think that that is the solution here, and I don't think it explains the differences that we've seen.
Yeah. I think that's particularly notable given we've seen some papers that highlighted a much higher use of steroids, some upwards of seven-plus times that amount that we used in iMMagine-1, and they still failed to prevent the development of delayed neurotoxicities with Carvykti. One, as Chris highlighted, mechanistically, it doesn't make any sense. Two, empirically, it doesn't make any sense. Three, as Chris also spoke to earlier, we're seeing pretty consistent rates of delayed neurotoxicities across Parkinsonism, cranial nerve palsies, and enterocolitis in the real world in the FAERS database quarter after quarter despite these interventions. I think that all points to something that is specific to the mechanism of action of a particular drug rather than something that's modifiable by steroids.
Okay. That's very clear. Just to be crystal clear here, for the EHA presentation, will it have the same cutoff as what was in the release? I believe it was May 1st, if I'm not mistaken, or will it be a little bit later?
No, that would be the—go ahead, Chris.
Yeah, it's the same data cutoff.
Okay. All right. Obviously, there will be another iMMagine-1 update at ASH in December, presumably. I guess as we think about that update before the end of the year, it's going to be very similar, I guess, with just a few months later follow-up, maybe, I guess, five to six months more follow-up. Is that the right way to think about it?
Yeah, that's right. I mean, typically, we've had data cuts sometime in the October timeframe for ASH, just mechanically how long it takes to do the data cut and review and align with our partners and all of those sorts of things so that we would give you about another five months of follow-up.
Okay. Are you guys still on track for a BLA filing by year-end, potentially Q4? Is there any reason why you couldn't file earlier in the second half, say, Q3? Have you had all necessary FDA conversations?
Nice try, Tyler. We have not guided to specifically when the BLA filing will be. What we continue to guide to, and I'm happy to reiterate, is that we expect the launch somewhere between mid to late 2026. We are all working feverishly to make the closer to mid-2026 case happen. In fact, that would be sort of our hope and internal drive along with our partners at Kite. We haven't guided specifically to when the BLA filing would be. We are working as hard as we can to move as quickly as we can. I think we can update that we have had very productive meetings with the agency, not just on anito-cel, but across our pipeline in the first quarter of this year. We feel like we have the clarity to move forward with the file.
Fair enough. Mid definitely sounds better, mid-2026 than late 2026. Are you guys still confident in the potential to have an accelerated review process?
Yeah. We do expect to still have a priority review. We do still expect to have full approval based on precedent as well as the strength of the data. I think maybe I'll jump ahead to, I'm sure, a question you're going to ask with respect to the new appointment at CBER. Look, I think I will say two things. I think one, on the margin in these large organizations, whether it's the FDA or a large company, things you don't tend to see changes. But on the margin, we're actually quite happy and pleased with the appointment of Dr. Prasad as someone who actually knows myeloma, who recognizes some of the challenges with the existing therapies, and was particularly focused on safety.
We feel like we're bringing a therapy to the agency and ultimately, hopefully, to the myeloma community that addresses a lot of the challenges that have existed to date with myeloma therapies, whether that is some of the safety issues we spend a lot of time talking about today, whether that's the scale and reliable manufacturing that's been a challenge to deliver in the myeloma community. We feel we've got a really great therapy here. We feel like it's backed by really strong evidence, and it certainly warrants both, in our view, as well as in Kite Pharma-Gilead's view, priority review and full approval.
That's great. Yeah. You think the lack of delayed neurotox alone might be meaningful as per the New England Journal Medicine publication by Dr. Prasad. Maybe on manufacturing, transfer has been complete, vein to vein time is within the high end of the range of the commercial products, which is 14-17 days, even though they've said they moved to 14 days more recently. Is there anything left to complete on the manufacturing front prior to regulatory filing or approval?
Yeah. I mean, I think that's a really important point that I feel like is often overlooked, Tyler. One of the challenges for both the prior BCMA CAR-T launches were manufacturing-related. One was on the vector side. One was on the cell processing and manufacturing side. To have the—we're not switching vector. We're using the same vector we used in iMMagine-3, so that ameliorates that issue. The tech transfer and manufacturing has already been done by Kite, as you highlighted. iMMagine-3, Kite continues to deliver just incredibly well right out of the gate with 100% success with a manufacturing time, as you said, within their commercial spec, which is under 17 days.
We feel really great about where we are and Kite's ability to deliver this at scale, at launch, and not have the sort of challenges in the ramp that others have had to endure.
Okay. Maybe just a follow-up on that. As we look forward to the launch next year, hopefully mid, but certainly by the end of next year, the process is the product here, right, which is why manufacturing is so key. How do you view the launch progressing relative to what we saw with Carvykti, especially as we think about it potentially being fourth line plus versus fifth line plus?
Different. We expect the launch to be different in a good way. When we actually announced this deal, so going back to late 2022, the thing that we resoundingly heard from the physician community is that the myeloma physicians wanted the same experience as our lymphoma colleagues in terms of the scale, availability, and reliability of therapy and of these therapies. We are excited to be able to bring that. When we talk about also availability, you have to factor in the commercial footprint that Kite has, the ability to deploy into that footprint very rapidly upon launch, combined with their manufacturing excellence and our combined efforts from a sales and marketing perspective.
We expect to do something that hasn't been done in BCMA CAR-Ts or, frankly, in CAR-T launches more broadly, just given the size of the myeloma market and deliver in a differentiated way that will hopefully delight our customers, save a lot of lives, and build a strong business.
Wonderful. In the last few minutes here, we're not going to get to it all, but maybe we can knock off at least a couple of client questions. Had someone asked if the improvements in efficacy or safety could be observed due to better prior lines of therapy. Curious to get your thoughts there.
I think, Chris, you can cover that. I think the question is whether lines of therapy have an impact on efficacy. The short answer is no, but Chris can get more color.
Yeah. I think, yeah, I think that's—I think there are a lot of things wrapped up in that conversation, and it is a challenging one to answer without data having been disclosed yet. I'll just say we have looked at our data from the phase one, and we've looked at the data available in the literature, even from other programs. I think what you can see actually quite clearly is that the patients who have had more prior lines of therapy actually tend to have better outcomes on average than patients who have had less prior lines of therapy in this relapse refractory setting.
The reason for that is most likely because when a patient goes on a study like this and they are entering fourth line versus entering fifth line versus sixth line, the patients in the fourth line may have run out of good therapeutic options faster, which may indicate that the biology of their disease is somewhat more challenging than someone who made it to sixth line despite not having all the currently approved agents at that time, right? Remember in the days when you were going into sixth line CAR-T in 2020, there was no talquetamab or tocilizumab or other options that might have slowed down disease progression.
Now, that is kind of the broader challenge I want to come back to, which is the reason why we believe that CAR-T will become a very important therapeutic option in second line and why iMMagine-3 matters so much is that what we are seeing now is an evolution to more of the therapies being given in first line in the combination of therapies given in first line. By the time a patient enters second line, they are oftentimes refractory to at least one, sometimes two, sometimes three agents, and they're going to need other therapies. Coming back to your original question and sort of the—I think the underlying philosophical question people are asking, we don't believe the literature suggests that in the relapsed refractory setting that fourth line patients do significantly better than fifth plus.
Obviously, we think that that will probably be supported by data in the future as well, although we do not have any to show right now. It is a bit of a nothing question in my mind because when we look at the things that actually predict outcomes in things like matching adjusted comparisons, those things are the refractoriness status, the disease burden, the presence or absence of extramedullary disease. Those are the things that really matter. It is not what line of therapy. In fact, that is from a publication that J&J put out. I think we can rely on that as something that is very indicative of what happens for those patients.
Okay. Maybe just one more quick one before I let you guys go since we're up on time related to ALC. Just a client was asking, since you've looked at ALC levels, even though you're not using it to treat patients, you've looked at ALC levels, you've seen a much cleaner safety profile. They're asking, is the ALC level data comparable to what Carvykti has disclosed? And therefore, if this suggests that the whole hypothesis that J&J Legend is putting out there is inaccurate in terms of ALC levels being above 3,000?
It's hard to tell because we've only seen spot data. Just the short answer on that, it's not like they've released all of the ALC data for CARTI21. I think if they did that, we're happy to take a look at it and give you a more detailed answer. Looking at handfuls of ALC data from series from this site or that site is not particularly informative.
Okay. Fair enough.
I'll also just add, I think it's also important to think about what ALC represents. It's a surrogate for CAR-T expansion. It's actually what links these events to actually Carvykti. Remember, these are different products. The ALC data for Carvykti is marking Carvykti expansion.
Okay. That's very clear. All right. Since we're up on time, we'll go ahead and wrap things up. Rami, Michelle, Chris, thank you very much for your time and the discussion. We look forward to the formal presentation here in a couple of weeks, and we'll talk to you all soon.
Thanks, Tyler.
Thanks, Tyler.
Thanks, everyone.