Hey everyone. We are back. Our next fireside chat is with Fred Aslan, CEO of Artiva Biotherapeutics. Maybe I'll kick it over to Fred for a quick overview of the company, and then we'll get into a Q&A. Fred, over to you.
Thanks, Alex, and thanks again for having us. I'm Fred. I'm CEO of Artiva. I've been CEO of Artiva for a little over five years. We are a phase III company developing a therapy for refractory RA. We are one of the companies that is pursuing the very exciting mechanism of deep B-cell depletion, as first demonstrated by autologous CAR- T, and we're doing it using NK cells. The advantages of NK cells is that we have demonstrated that it can drive a very deep B-cell depletion, and there are two specific advantages. Number one, we don't have the side effects that are typically seen with T- cells, such as the CRS and the ICANS. Number two, because of our differentiated safety profile, we can treat our patients in a community setting.
Everybody understands that for any of these deep B-cell depleting products, the holy grail is to be able to do it in a community setting where the vast majority of the patients are located. I can tell you that to date, we've dosed over 100 autoimmune patients, and the vast majority of them we did in a community rheumatology practice. We chose to work in RA because not a lot of people had been working in RA, and we wanted to go after an indication where we could be differentiated. It's a large indication. It's one where having a tolerability profile that's differentiated is important, and I'm sure we'll be talking a lot more about that as we speak.
Great. Well, thanks, Fred. Maybe to start, I'd love to dig in a little bit more on the AlloNK approach, the combination with rituximab, and mechanistically why that gives you some differentiation on the safety side in particular relative to call it an autologous CAR- T or a T-cell engager.
Yeah, definitely. The goal of the therapy is to eliminate all the B- cells. This is what we've learned through autologous CAR- T. That is the goal of the treatment. We use non-genetically modified NK cells because NK cells turn out to be equally strong effector cells as T- cells. Because we don't engineer our NK cells, so we don't engineer a CAR, we actually use monoclonal antibodies to do the targeting. Rituximab does the targeting towards CD20. Then the Fc portion of rituximab will activate our NK cells via the CD16 receptor. That activates the NK cell, which kills the B- cell. This is an entirely different approach to actually killing B- cells, but it's pretty much using the same CD19, CD20 targeting.
The big advantage of NK cells is the fact that they don't expand the way the T- cells have to expand in order for them to do their activity. It's that T- cell expansion that leads to CRS, ICANS, and some of the recent autologous CAR- T side effects that we've been hearing about from Novartis and from BMS . On the one hand, the doses of NK cells that one has to give is a couple of orders of magnitude higher than the T- cells. You have to have a very scalable manufacturing process in order to do that. If you can do that and you can already give them at high doses, then the advantage is you're not relying on them to expand, and so that avoids a lot of those side effects.
Those side effects become really important because as one is thinking about bringing these sorts of therapies into the community, whether you're thinking of taking autologous CAR- Ts, TCEs, in vivo CAR- T, or even our therapy, one ideally has a tolerability profile which allows a physician to actually treat their patients and send them home so that they fit into the existing monitoring profile that those have. As we'll talk about our safety data, thankfully our therapy does seem to have a very strong tolerability profile.
Are you using the normal dose of rituximab as your backbone as well, or do you have to have a higher dose for that?
No, we're using the exact same dose of rituximab that's in the label for RA. We use the same dose of rituximab that was in the label for NHL when we treated NHL patients with our therapy, and we achieved high complete response rates and really nice durability there. In autoimmune disease, we're using rituximab as rheumatologists use rituximab, which is basically two doses of a gram each 14 days apart. On those same days is when we give our AlloNK, which is just another 5 to 10 minute IV infusion.
Then maybe sort of at a mechanistic level, how does the NK cell-based depletion that you're doing with your approach compare across this systemic and also in the tissue compared to your classical CAR- T or T-cell engager approaches, at least from what we've seen in autoimmune so far?
Yeah, I think that the data that has been generated in oncology is very telling because you see our hypothesis in autoimmune disease is you have to get rid of every B-cell and then hopefully you see a benefit in immune disease. When you are treating NHL, it is a little bit more categorical than that because if you do not kill every B- cell, then your cancer is still present. What we demonstrated in 45 patients that we treated with a non-Hodgkin's lymphoma, and then another 24 with Hodgkin's lymphoma, is that we do eliminate all the B- cells using this approach. It is a different cell. It relies on a monoclonal for targeting, whereas the other genetically engineered products rely on engineering in a CAR. But the ultimate effect, which is this deep B-cell depletion in both the periphery and the tissue, has been demonstrated by our therapy.
Okay. Let us walk through the treatment process for a patient. Maybe to start in the clinical trial setting, you have a patient, what is the process to get the drug?
Yeah. We have validated this process now across dozens of sites, and as I mentioned, we have treated over 100 autoimmune patients, which speaks to how relatively easy it has been for community practices to enroll patients and to treat them. So you go through screening, and if you pass screening, you will get scheduled to come in on a Wednesday, Thursday, Friday to get your cyclophosphamide and fludarabine. So you have those, and you go home, and then you come in that next Monday to get your first dose of rituximab and to get your first dose of the cells. And then you come back two weeks later to actually have your second and final dose of the rituximab and the cells. So five visits to the IV infusion center, and rheumatologists are used to giving rituximab. They are used to giving cyclophosphamide.
Giving fludarabine is no different than giving cyclophosphamide, and then our cells themselves is a 5 to 10-minute IV.
Why do you need preconditioning in this context?
It hasn't been proven yet in the field of cell therapy that cell therapy works without conditioning. There has been extensive work done in the context of oncology to try different cyclophosphamide and fludarabine regimens. If you read all the papers that were published at the time, this is in the 2012 - 2015 era, it was pretty determined that without the cyclophosphamide and fludarabine, the autologous CAR- T did not work as well. When you translate that into autoimmunity, the question is whether the same is needed in autoimmunity. We've had a couple of experiments of allogeneic cell therapies use, for example, Cy only and no Flu. When they did that, they later came out and suggested that when they only used Cy, they did not achieve a deep level of B-cell depletion, and they decided to embrace the Cy and the Flu.
We have a couple of efforts out there from companies that are trying not to use cyclophosphamide and fludarabine. I would tell you that exploratory work is very early and hasn't convinced the community yet that you don't need cyclophosphamide and fludarabine. What I would tell you is we are all very interested in understanding what is that lowest level of cyclophosphamide and fludarabine that you need to use.
Yeah.
But in the meantime, like I told you, we have 100 patients that we have treated in a community setting. Hospitalization has been low. No CRS, no ICANS. We're starting with something that already looks pretty compelling, and over time, we can look to optimize it. Now, to get to your specific question, why is it used? The main reason it's used is because you want to give the host an opportunity to let this allogeneic NK cell do its thing during that couple of weeks. That's all you're looking for, a short window of time when the host cells are not going to attack and reject the cells that you're giving them so that they can get rid of all the B- cells. And then literally two, three weeks later, we see the cells coming back.
And during that period that you are cytopenic, you're using prophylactic anti-infectives. You have antivirals, antibacterials, antifungals, and that explains why the infection rates are so low and why our hospitalization rates are so low.
Yep. Makes sense. And maybe I want to transition to your clinical data that you generated so far, but I also wanted to level set around, because I think there's a lot of work being done right now in, quote, "refractory RA populations," and I think there's no, in my opinion, clear definition of what it is that people are talking about. So when you're talking about refractory RA, what does that mean to you relative to some of the classical CAR- T experiments and then some of these other more specific, difficult to treat populations that folks are looking at?
Yeah. So in RA, the Europeans have actually established a term called difficult to treat RA. The main backbone of that definition is patients who have failed two distinct classes of biologics or targeted synthetics, failed at least two. There's a few other criteria that go into D2T, but D2T is not a definition in the U.S. In our conversations with FDA, when we discussed what refractory RA is, we leaned on this concept of a patient that has failed two or more distinct biologics or a JAK.
What you find when you look at the ACR50s of the approved products is that, first of all the products that have been approved in RA, and there are many of them, we can talk about that, they have all been approved either as your very first biologic when you have failed methotrexate or your very first JAK, or they have been approved in what's called a TNF inadequate response population, which is basically you failed TNF, now you're going into your second drug. What we are talking about is when you're going into your third drug now, you've already maybe tried a TNF, you've already maybe tried a JAK, and now you remain in high disease activity, and you have no great solution. You see the labels from the products, if you're taking one of these agents for the first time, is actually quite high.
You can achieve ACR50s in the 40% range. By the time you're using it in a TNF failure population, those ACR50s, they decline into the 30% range. When you look at what's available in the literature, because there is no product with a label that specifically discusses patients that are refractory, but we know what that looks like from the literature as well as from real-world registries, we see that the ACR50s are in the 10%-20%. What we are targeting is a greater than 50% ACR50 in this refractory population, where the standard of care is really in the 10%-20%.
Yep. That's great. I guess sort of in that context, can you talk about the data you've generated so far in sort of this early proof of concept work you've done in RA and sort of how that translates to how you're thinking about phase III?
Yeah. We shared, for the first time, our clinical efficacy data in autoimmune disease at this past EULAR. There we had data on roughly 35 patients with roughly 20 of them in RA. The numbers are relatively small, and they continue to grow because we continue to enroll in the basket study. As I mentioned, now we have over 100 patients that we have treated overall with autoimmune disease. We saw an ACR50 of 71% in that population that we presented. Again, these are small numbers, and our target is achieving an ACR50 of 50%.
The fact that these patients were very refractory, and they were very much the types of patients that we are looking to enroll, it gives us confidence that once we read out on our randomized control trial, which should read out in a couple of years, that we should see the targets that we're looking for.
Yeah.
Now, we didn't just share data in RA. We shared data in Sjogren's as well as scleroderma. These were the patients that were most advanced. Part of the reason for sharing data across indications, even though in each one of them, we only had between 5 and 20 patients, is because we wanted to show that the therapy is very active no matter what indication you're using it for. We had already demonstrated in multiple B-cell cancers that it was very active, and now by putting forth data in three distinct indications and showing autologous CAR T-like indication, we're trying to demonstrate a consistency in the efficacy of the product.
In terms of length of follow-up and depth of the response, can you sort of give us a little sense of kind of where you're at at this point in terms of the RA and other data sets?
Yeah. With the follow-up that we presented at EULAR, we had three patients that had reached the 12-month mark with RA, and they had not had the need to go on another therapy, right? The durability that we saw on that cut was at least 12 months. It's important for me to note at this point, as you ask about durability, that one of the big differences between our therapy and autologous CAR- T, for example, is that redosing is part of the paradigm. The way we think of our target product profile is that you treat a patient, the patient experiences these long periods where they benefit from the drug, and then they can be redosed again.
We demonstrated in oncology that we can redose and that we achieve similar PK and PD, and we still need to demonstrate that in autoimmune disease, but the oncology data gives us confidence we can do that. Our target, as our N gets bigger and as the follow-up gets longer, is to have the vast majority of patients that have a response, that they have at least a 12-month of durability. Because then if you imagine a therapy that you are using to retreat every 12, 18 months, and you're keeping a patient in a response for years, that's how we can measure durability as opposed to just a single treatment.
And then you mentioned no CRS or ICANS. Infection profile looks solid. I guess, is some of the neutropenia you're seeing, is that a function of the Cy/Flu, or is that a on-target effect as well?
It's mostly a function of the Cy/Flu, because when you give Cy/Flu, you do cause a transient, and I would say shallow, cytopenia, including neutropenia. But, a couple of things that I would point out. Cyclophosphamide is not a very popular agent, right? If you just ask physicians, "What are your thoughts on cyclophosphamide?" If you ask them in a vacuum, they're going to say, "Well, I don't really like cyclophosphamide," right? But we have to remember a couple of things. We have to think about the target product profile as the whole thing, because the efficacy we are achieving, we are achieving in the context of the full regimen. Nobody ever says, "Oh, I love this drug, but if you could only not use what's causing liver failure, that would be great.
You have to think about the whole package. Cyclophosphamide, the doses of cyclophosphamide that are typically used by rheumatologists or have been used in the past by rheumatologists, are either high doses in the context of, for example, lupus nephritis, or they are doses which are similar to the doses we are using, but they use them over months. The low dose cyclophosphamide guidelines, EULAR lupus asks you to keep patients on cyclophosphamide over a three-month period. Because of these high doses, and because of this extended use, cyclophosphamide has a reputation for actually causing infections.
But because we use a low dose and we use it once, and that cytopenia is limited to a few weeks, we are not actually seeing that. In our first 55 patients, there were 2 patients that were actually hospitalized in the first 28 days, and they were not drug-related. One was a diabetic that had diabetic ketoacidosis, and the other was a Sjogren's patient that experienced dehydration. We are not seeing those high-grade infections that physicians are concerned with. To answer your question, yes, the neutropenia and the overall cytopenia that you see is driven by the Cy/Flu, but because of the doses, it's transient and it's very shallow.
Yep. Okay. That makes sense. I would love to sort of walk through a little bit your discussions with FDA and how you got to the single phase III trial and kind of what that looks like.
Yeah. Based on the conversations that autologous CAR- Ts have had with the FDA, we approached it from the point of view of the FDA has been quite flexible in terms of how it thinks about the data that needs to be generated for approval. We figured all along that it was important for us to have an active arm, because if we didn't have an active arm in our registrational trial, and we came out with an ACR50 greater than 50%, I could see either the investor community or the physician community being like, "That looks great, but I wonder what rituximab would have done on its own in this population."
We figured, let's randomize patients to either rituximab or to receive our therapy, because then we will have a definitive experiment showing how much better, if it works, our therapy is compared to rituximab, which works as well as any agent in this refractory population. Because we are using ACR50 at six months as the primary endpoint, we offered a 150 patient study, which would be sufficient to both demonstrate this delta in signal that we are looking for, as well as to provide the safety database for the product.
The agency agreed, number one, on a single registrational trial. They agreed with the statistical plan and the number of patients that we are collecting. In terms of the safety database, they would take all of the patients that we treat in this phase III, along with all of the autoimmune patients that we will have treated outside of this RCT, and as I mentioned, we've already treated over 100 patients outside of the RCT, which then form the basis of the safety package that we would be presenting to them at the time of the BLA filing.
Yep. Makes sense. I guess, as you think about ultimately this drug succeeding and potentially getting approved, I think that it's straightforward to consider your bar on ACR50 from that perspective. How about considering the dosing paradigm? How is FDA thinking about what they need to see to say what is the indication statement to what does the dosing look like in terms of duration, follow-up, all of that sort of thing?
Well, the dosing, we're using pretty standard dosing. We have used slightly different doses. We've kept the dose of rituximab the same because that is the dose that's approved in the label. cyclophosphamide and fludarabine, we tried slightly different doses, and they all seem to be doing what we need them to do. The dose that we are using for the RCT is pretty standard dosing, low dose 300 mg of cyclophosphamide per day with 30 mg per meter squared of fludarabine. It is low dose cyclophosphamide and fludarabine. Then we are using 4 billion NK cells per dose.
A total of eight. We have tested three, and we have seen good activity. We have tested 12, and we have seen good activity. In the RCT, we are working with eight to make sure that patients are getting really good exposure and maximizing the chances that they are doing well.
Yeah. And I guess I mean.
The FDA agreed that with a single dose. Sorry?
I guess I mean more around redosing. What do they need to understand about your redosing on the label? What data will you have around recommendation for when to redose patients?
Oh, I see. Yeah. What we're doing is we're implementing a mechanism to take patients that have responded to the drug, then once they lose their response, we're creating a mechanism for several of those patients to be retreated. Because then we can look at what is the PK of the AlloNK when you are retreating. What is the PD? Can you actually drive B- cells to the limit of detection the second time around? If you can, it means that your drug is not getting rejected, right, by the host's immune system. Then can you drive clinical benefits on that patient upon retreatment that looks very similar to what it was to begin with? So that is data that we are collecting, not on the RCT patients, but on the patients that we are already treating, so that the time matches.
By the time we're going to FDA for the BLA filing with the data from the randomized control trial, we also have some initial retreatment data that we can also take to them so that they can see how it works.
Yep. That makes sense. Obviously, refractory RA is not a small number of patients, and your ability to sort of be in the community could potentially allow you to sort of treat a lot of patients. I guess you've alluded to your manufacturing process as a key part of the whole story here. Can you talk about what that looks like and what are the COGS associated with a product like this?
Yeah. So very importantly, that was always our biggest differentiator compared to most others in the allogeneic cell therapy field. Artiva was founded in 2019, but it was a spinoff of a Korean company, GC Cell, that had already, the decade prior, developed a scalable process. So starting with our very first patient in oncology that we first treated back in 2021, we have been using the same backbone process, which is a very scalable process. We source umbilical cord units. From one umbilical cord unit, we can make enough therapy to treat over 500 patients. As we have discussed in the past, at scale, this product has the opportunity to have COGS in the $8,000 range for an entire treatment. So that's 8 billion cells that are being given to patients.
As you appreciate, that's more than an order of magnitude lower than what autologous CAR- T can do. That's more in line with what biologics can do. But very importantly, I personally think it's not possible for allogeneic cell therapies to become so cheap because they are mostly using genetic engineering. When you have to genetically engineer the cell, that adds two specific issues. One, it's expensive to genetically engineer cells and to do it scalably. The second is, once you genetically engineer the cells, it's expensive to run clinical trials because you have to follow these patients forever because of the risks of insertional mutagenesis, which is a lingering risk that we just don't carry because we don't genetically engineer them.
We have a 9,000 sq ft facility, 9,000 in our San Diego office, and that is enough for us to treat at capacity over 500 patients. A very scalable process. We do not use CDMOs, and our burn is fairly low because we are not spending as much as our peers are in trying to make drug product.
Makes sense. Then maybe final question here, can you talk about your current cash runway and what the embedded assumptions are here?
Yeah. We raised $300 million in May of this year. This carries us into 2029. As I mentioned, our randomized control trial, we are guiding that we would have a readout of that before the end of 2028. So that cash should take us through that readout and through sharing data across all the indications, data on retreatment, data on PD upon retreatment, and we will keep people tuned on other developments.
Great. Well, Fred, thank you so much for joining us today.
Thanks for having us, Alex.