Excellent. All right. Well, good afternoon, everyone. My name is Jeff La Rosa. I'm an Equity Research VP on the immuno-oncology team here at Leerink Partners, and I'm excited to welcome Senti Biosciences for the first full day of our annual Global Healthcare Conference. You know, joining me from Senti to participate in a fireside chat is Chief Executive and Co-Founder, Timothy Lu, you know, CMO, Kanya Rajangam, and CFO, Jay Cross. Thank you all for being here.
Thanks for having us.
Let's kind of dive in here. I guess, Tim, just to sort of set the stage here, could you maybe just give a brief kind of overview of Senti? I know you guys are focused on logic gating and gene circuits, and have really kind of pioneered those technologies. You know, what really is that in the context of cell therapy and what could, you know, therapeutic challenges does that solve?
Yeah. No, thanks a lot for the opportunity to be here, guys. What we're trying to do at Senti is solve one of the key challenges, I think, facing oncology drug development. One of the big areas that everyone's trying to do is how do you kill cancer cells without killing the rest of the patient, right? Traditionally, the strategy for doing that, whether you're talking about a T-cell engager, an ADC or a cell therapy, is you try to find some sort of target that has that therapeutic window that gives you that clean cancer targeting or results in, you know, some sort of side effects that are manageable enough.
It turns out there's only so many clean targets that we know about out there that are on the cancer cells that you can kill without, you know, sort of affecting too much of the rest of the patient. What we've been trying to do is go after the other cancers that are not in that category. What are the cancers where you don't have a clean target, where if you simply go after a single target that might be there, you're gonna cause, you know, such toxicity that you can't get the efficacy you want. That's where the idea of logic gates comes in. It's basically thinking about how do you recognize cancer cells more accurately and kill them, but also how do you recognize the key healthy cells and protect them from being killed.
We think cell therapies are uniquely designed to do that. You know, cells have this ability to float around in the body, sense what's going on, and make decisions at a level where, you know, a small molecule or protein therapeutic can't. We have spent a lot of time designing cells that can essentially recognize multiple targets inside of the body. If the targets are on the cancer cells, they basically trigger killing, and if they're on the healthy cells, they trigger protection. The way we actually do this in reality is we basically are designing cell therapies with multiple CAR receptors. The activating CAR is the thing that recognizes and kill cancer cells, and then the inhibitory CAR is the thing that recognizes and protects healthy cells.
We have a CAR that turns on and a CAR that turns off. This comes down to a lot of genetic engineering expertise we've built in-house. We've had to figure out how to make really good activating CARs and then pair them up with really good inhibitory CARs so that you get this sort of balanced behavior. We're excited about this in that, you know, we've shown this works in both NK cells and T cells. It's a general, you know, sort of engineering that we can put into any cell type. Last year we showed a lot of proof of concept in humans, both in terms of top-line clinical data, but also, you know, much more detailed analysis of patient samples showing that essentially we're killing cancer cells and sparing healthy ones.
Yeah, let's definitely dive into that first program, SENTI-202. Before we get into the data, could you sort of talk about the specific logic gates that you employed in that construct, and what challenges of AML specifically do each of those solve?
Yeah. I think when we look at indications, we're really trying to find diseases where we're not really going up head to head with any other existing T-cell engager, ADC or cell therapy, right? We don't wanna go compete in areas where, you know, like, there's 100 different programs going after the same target. We were looking for indications where you can't find that clean target, and AML stood out to us as one of the low-hanging fruit in that people have been trying in AML for a very long time to identify targets that could cleanly kill cancer cells but have enough therapeutic window to work. Now I think in the clinical literature, there's been multiple pieces of evidence showing that such targets are basically super hard to find or if they're there, they're not very generalizable.
In AML specifically, we've designed a cell therapy that recognizes two AML targets, and that's CD33 and FLT3. Both these targets are well known in the space. You know, CD33 is expressed on AML blasts. FLT3 is expressed on AML leukemic stem cells. With a single activating CAR, we can recognize either of those targets and trigger killing. The problem with these two targets, as well as many other AML targets, is they're also expressed on healthy cells, for example, healthy hematopoietic stem cells. What we did was design a inhibitory CAR that recognizes those healthy cells through a target called VSIG2. And basically when it recognizes that target, it protects those healthy hematopoietic stem cells from being killed.
We identified that VSIG2 safety target based on our own bioinformatics analysis and basically showed that VSIG2 is high on healthy cells, healthy hematopoietic stem cells, but low on cancer cells. It has the opposite expression pattern of almost all cancer antigens that people are looking for. It's through this sort of, like, multi-specific targeting where we can get this activity. That's our program. We have an allogeneic program that leverages that logic gating and we'll tell you more about that later.
Yeah. I think part of that is not just the logic gating, but if, you know, the cell chassis is an NK cell.
Sure.
It's allogeneic or off the shelf.
Yeah.
You know, I think some people have preconceptions on what an NK cell.
Sure.
M ight have limitations, et cetera. You know, it's in AML where patients are frail. You know, there's been struggles trying to get autologous T cell or T-cell engagers to sort of show efficacy in that setting. I mean, is this not a liability, but a strength to have an NK cell? What are your kind of perspectives on that?
Yeah. I think coming into it for us as a company, we don't have preconceived notions that we have to pick a certain cell type. For us it was, okay, for a given disease type and what the logic gates can do, like what's the right cell type for the patient, right? Because our technology works across many different types of immune cells. In AML specifically, we think NK's make a lot of sense. These patients are frail. It's hard to get good quality autologous T cells from these patients. That puts a limitation on CAR T. Two is an allogeneic product has some advantages here in that relapsed refractory AML patients have a unfortunately very short time to live, like a median survival of maybe five months or so.
If you're gonna waste a like a month or two for vein-t o- vein manufacturing, that's another bottleneck which you'd like to avoid. Three, there's a lot of literature around sort of NK cells having some natural anti-AML killing activity. All of these sort of are in that pros category where we felt that an allogeneic NK chassis would make sense for a disease like AML. That's why we went forward with SENTI-202 in that context. There are some additional pipeline programs in the solid tumor space where one may go through the same analysis and decide that a T cell makes more sense, and we've therefore developed our technology to be generalizable in that sense.
Absolutely.
Yeah.
Going into data, I mean, you at ASH, you had pretty encouraging update, phase I update. I guess to sort of start, I mean, just maybe summarize the key takeaways and findings that you saw from that and how that kind of data set evolved from the AACR update originally and kind of where you left 2025 entering this year in terms of the clinical data.
I'll let Kanya go into the details, but I would say at a high level, we want to show in this phase I study a couple things. One is, we want to obviously look at the, top line responses in these patients, the efficacy, the safety, et cetera. Two is we really want to dig in mechanistically, given that this is a new technology that really hasn't been proven out yet in humans, and get patient samples where we could say, "Are we killing the cells we're trying to kill, and are we sparing the cells we're trying to spare?" We've been able to achieve both of those goals, and maybe I'll let Kanya talk about that.
It was a terrific data presentation for us to have at ASH. We had an oral with our phase I 20 patients. We also had a correlative poster where we were able to look at how our logic gates are working in patient bone marrow samples. In our 20- patient data set, what we saw, first of all, the main goal of a phase I is to identify a go-forward dose and regimen, which we did. We were able to start at a very high dose because of the excellent safety profile of NK cells. We made the case that we wanted to start at a dose that was biologically active. Some small dose finding and then, you know, much of our 20 patients got our go forward dose.
We saw really good, deep, durable responses in relapsed refractory AML patients. We had a 42% CR/CRh rate in our recommended phase II dose. 100% of our CR patients were MRD negative and a median durability of 7.6 months and counting. We are very pleased with our results. On the safety side, consistent with the NK safety profile, you know, SENTI-202 related events were Grade 1 or 2 fever, delayed infusion related reactions on the day of dosing. Very well-tolerated. In terms of our correlative, just quickly to touch on that, one of the things we did was we do CyTOF on bone marrow samples for pre and post. That's a way you can look at the different cell populations in the bone marrow.
We saw that patients who were treated with SENTI-202, their leukemic stem cells, which is one of our key targets, start to drive, you know, short duration of remission. Those are eliminated in our patients who responded. Importantly, going back to the logic gate design, our hematopoietic stem and progenitor cells, these are the ones necessary to repopulate the bone marrow with healthy cells. Those were protected or enhanced in all our patients achieved a CR or CRh. Good validation of our proof of concept and very concordant with the clinical data as well, where we saw very rapid count recovery.
Yeah. Could you put into context actually the composite CR rates and the durability and preliminary survival that you showed, you know, relative to what these patients might get in a normal treatment setting?
Yeah. Yeah. The relapsed refractory AML patient population is a very high unmet need. The median survival for these patients, a large retrospective study showed 5.3 months was the median survival. There really isn't very much for these patients. There is a handful of drugs approved for mutationally selected targets, most notably recently with the menin inhibitors for a subset of patient population who have the right mutation. Even there, the CR rate was about 12%. CR/CRh together in the 20s, about four to six months median durability. Those who don't have those mutations get salvage or recycled chemotherapy that they've already seen and failed in the front line, which is why they're now relapsed refractory. It really is a huge unmet need for these patients.
Seeing the kind of results we've seen so early in our phase I is very encouraging for us, you know. What we are really focused on now is taking it to the next phase of development. Yeah.
Yeah, sometimes I get questions, you know, given you used fludarabine/cytarabine conditioning.
Yeah
And contribution of components, you know, often comes up in this type of context. I know you had some correlative data around this. If you could just speak to, you know, why you're confident that the signal that you're showing is above and beyond what you'd expect from the fludarabine/cytarabine conditioning.
Yeah. I mean, the data tells us, right? Fludarabine/cytarabine has been around since like the 70s. I don't know, even before I went to medical school. We still have a big need for AML patients here. It's clearly not something that has, you know, helped these patients in any way in its current form. Two large, prospectively, like large multicenter trials recently looked at a variety of chemotherapy combination arms. One was the ADMIRAL study for FLT3-mutated patients where gilteritinib went against standard of care. Another one was an experimental agent which didn't work, but also looked at seven different standard of care options. The CR rates in both of them, 12% in one, 10% in another, median survival of three months, right? That's not what we are seeing. That's one.
Secondly, even within our patient population, we have had patients who've come on fail fludarabine/cytarabine in combination with other chemotherapy agents, it's often given in combination with anthracyclines and other things, come on our trial and then responded to SENTI-202. That gives us confidence in this. One of the reasons we chose fludarabine/cytarabine was actually there was two reasons. One is we did want some debulking, and because the way NK cells, what we are designing it to do is to go and eliminate the last AML and LSC. The second reason was pre-clinically what we found was there's a potential for synergy because when you expose AML cell lines to cytarabine, it causes upregulation of our targets. It becomes a one-two punch. Yeah.
Yeah. I think another thing that was pretty striking in the dataset was the safety profile was quite encouraging. CRS of course, comes top of mind with cell therapies, but, it's, you know, it, you know, quite potentially more important is, you know, non-relapse fatality from things like infections and severe infections 'cause you're giving really myelotoxic agents, oftentimes with these sort of, you know, more intensive chemotherapy regimens, but it seemed quite low. Was there something in the hematopoietic stem cell, progenitor recovery that you saw that you think kind of was supportive of the knock gate potentially? Or maybe it's the NK cell, or if you can just kinda speak to that context that might give you confidence in its path forward.
Yeah, no, that's a great question. One is the inherent safety of NK cells extremely well- tolerated. All we saw was some delayed infusion related reactions. It's small patient numbers in early days, but it also seems to be that the general hematologic toxicity and the side effects were also numerically lower on our study. I mean, it is small numbers, I'm the first to say that. What I will say is this is consistent with what we saw in our correlative data and the way we've designed this product. We designed this product to selectively kill AML cells and protect the healthy hematopoietic stem cells. What we found in our CyTOF studies is patients who responded, right?
Among responders you have CR, CRh, and those who are familiar with AML also know there's a category of response called morphological leukemia-free state, where AML is eliminated, but you don't have normal count recovery. When we looked at our responders, those who entered the trial with any level of HSPC at baseline, they all went on to get a full CR or CRh. We had a couple patients who came on with zero HSPCs at baseline. They're the ones who achieved an MLFS. That's again concordant data. If you look at our CR, CRh patients, by day 14 their counts have come back. Even we had, you know, a few patients who had some febrile neutropenia, thrombocytopenia, that's at the nadir of when with the chemotherapy conditioning everything goes down.
It's limited for a few days and then it recovers very quickly. That's why so far our rates of pneumonia, sepsis, so on, just three patients have had one of those events. We will see. We'll continue monitoring that, but it certainly seems to indicate that our logic gate is doing what it's supposed to do, in terms of not only eliminating AML cells, but protecting the healthy cells and allowing that bone marrow recovery.
Another thing, you know, that I noticed in the dataset was, you know, a high percentage of patients bridge to transplant, and of course that's a good thing in AML. That's often the treatment goal because it can be curative and has a really positive effect on overall survival. You know, it also happens to—t hat often ends up being the time point at which you are unable to potentially assess durability of the product on its own.
How do you sort of weigh what you're seeing in terms of bridging the transplant in terms of, you know, what the therapeutic potential is? Is that the goal of SENTI-202? Do you think patients, there's an opportunity to not need to bridge patients to transplant and sort of how are you thinking about that?
Yeah. The first thing I wanna say is our trial is not a bridge to transplant trial. In other words, our eligibility doesn't say that you need to be eligible for transplant, you need to have a donor. None of that, right? It's all comers, they can come in. What is the current standard of care for patients with relapsed refractory AML, especially ones who achieve MRD- negative CRs like they are on our study, it is to consolidate with transplant, right? That's what we are seeing. We do have a few patients who've responded who for other reasons did not get a transplant, and they've had very nice, clinically meaningful durability as well. I think what's gonna happen is as we put on more patients, we are automatically going to be generating data on both.
It's obviously very early days still. That is the standard of care and that's the right thing for the patient. I think we need a lot more data to be able to answer the question, do you even need to further consolidate with transplant or not in the setting of a therapy which is giving you not only MRD negativity, but also really decreasing your LSCs, because that's sort of the root of the cancer. That's gonna just take more data. Yeah.
You did recently announce that you finished completing enrollment in the phase I.
We did. Yeah.
What else did you need to see when you sort of rounded that out, and what are the next steps from here? Yeah.
We finished our phase I. We put on 22 patients in our phase I, and it's very consistent with what we had 20 patients data in ASH, very consistent as we have wrapped up enrollment into our phase I. Next steps now for us is to. One of the things that we, a big highlight for us from 2025 is we were given RMAT designation. That's an advanced therapy designation from the FDA, which allows us to have a lot more interactions with them even earlier in our clinical development program to essentially partner with them to make sure that when we go up to with the BLA, hopefully there are no surprises, right?
That is the purpose of the RMAT program, is to ensure that therapies which have some exciting clinical data, you can work closely with the FDA so that, you know, the finish is smooth. We are in the process of meeting with the FDA to discuss plans for our late phase trials now. That is the next step for us.
Yeah. What, you know, before, you know, this kind of, what's your preliminary thinking on what a registrational trial might look like given the profile you're seeing with SENTI-202? You know, the need to give, you know, fludarabine/cytarabine. I guess first, I mean, do you think with this type of relapsed refractory setting, I mean, is a single-arm trial sufficient for a full approval here?
One of the things that historically the FDA pays a lot of attention to is regulatory precedence. Seven drugs have been approved for relapsed refractory AML, all the different mutationally targeted ones, the two menin inhibitors, then there's an IDH 1 and 2, and FLT3 and so on. All of them used a single- arm trial to get a traditional approval. Gilteritinib trial is the one where they got their first label based on the CR/CRh rate, and then they came in with the survival from their ongoing randomized trial. We are very confident that that is the path forward, mainly because if you look at what's available in standard of care, I just referenced two large trials that were done with a CR rate of 10% or 12%.
You know, for us to even contemplate doing a randomized control trial, one of the things to consider is what would be an appropriate control arm. We are not really seeing that. That's why I think we, our plans are to go forward with that single- arm trial.
How many patients do you think that might require? Like give or 100, give or take?
You know, I'm not, we've not discussed that. Y ou know, we don't wanna front end, any discussions we have with our FDA coming up as well. You know, if you look at the literature, it has been in that ballpark. Especially with an effect size that we are seeing, it should be, you know, in that range. Yeah.
Yeah. In terms of like the type of patient setting, how does this play from like a fit or unfit perspective, which kind of roughly cuts the market in half? Would you sort of try to target specifically one, or is it kind of not even necessarily very clear when the unfit kind of funnels into the fit and patients will get venetoclax and azacitidine and everything, and then they will still get and prior intensive chemo? Does it all kinda come together here, or how do we kinda think about the opportunity?
Yeah. One of the big considerations for any AML therapy is the fitness of the patient overall. I mean, it is a disease of the elderly. The median age at diagnosis is 65, but there's, you know, like 75+ patients do have AML. So the way we think about it, obviously SENTI-202 is extremely well-tolerated. Everything in terms of who can come on our trial is really dependent on who are the patients who can get our low-dose chemotherapy.
Yeah.
That drives what, you know, who can come on. I would say we are, so we would certainly be skewing a little bit on the fitter side because they do have to get the full dose of fludarabine/cytarabine. It's not necessarily one wherein you can automatically say patients who are of this. It really becomes individualized. I would say we are right in the middle and skewing towards the fitter population.
Does biomarker selection matter at all with like your targets?
Yeah
or is it an agnostic sort of population?
Yeah, no, great question. We have, our targets are CD33 and FLT3. Ninety percent of AML is CD33 positive.
We do require our patients currently to come on with being CD33 positive as assessed locally. We don't have a minimal, none of that. When we did retrospective analyses, we saw that there was no real correlation, that a certain level of CD33 was needed. We will continue to monitor that. Obviously, it's 22 patients now, we'll see where we land as we put on more patients.
Yeah. As we kinda wait for this update, regulatory update, I mean, you talked to the FDA, do you so far planning to give another clinical update, maybe with more durability, and follow-up from the patients you've treated so far?
We've not announced when.
Okay.
Yeah, I, yeah, I'm sure that that's something we would do. What I will say, this is something we have said publicly as well. We presented 20 patients data at ASH. We now have 22.
Yeah.
It's more largely consistent. We haven't announced exactly when we will give our next update.
Yeah. The thing of like clinical supply, manufacturing, what's your readiness there and supply you have to treat more patients without further investment?
The last year, I know our clinical data gets a lot of play, but internally, we've also been really amping up and getting ready for our CMC with pivotal readiness. We've made tremendous progress in having our pivotal GMP process ready to go. That will also be a part of what we discuss with the FDA in terms of our approach, how we are thinking about it. We feel pretty confident, you know, that assuming everything goes well and so on, that we could be starting pivotal trials later this year. I think that's what we've said publicly.
As you kind of think about getting ready for that trial, I know you're plugged into the AML community, like what are investigators saying about this and their level of interest in terms of like potentially wanting to participate in a trial like this?
Yeah. A great question. We have tremendous investigator interest. One is our data, and two is unfortunately there's just such a lack of other alternative options for these patients. We have, I think pretty much every major AML leukemia center has reached out to us and wants to join us in our pivotal trial. Once our ASH data, once we had our oral at ASH, I think the next day they were like at least another 12 or 15 folks reached out because they wanna join in our next phase. A lot of excitement there. On that aspect, I think there's been, it's been received very, very positively.
Not just for relapsed/refractory, it's also interest in pediatric, and what's the newly diagnosed plans, e t cetera. I think all that is came out pretty clearly from the clinical community.
Yeah. Do you have any strategies to sort of maybe cost efficiently expand sort of the data for, you know, that you have, whether it's investigator sponsored trials, ways that, you know, other agencies could sort of help sort of fund different elements of the development here? I mean, what sort of strategic, you know, development or collaboration opportunities are you exploring? Yeah.
In terms of the collaborations, maybe Tim, you wanna address that? I can tell you that we are exploring. In terms of our clinical program, we are looking at not only, you know, looking at a pivotal for the relapsed/refractory AML, but ways to go earlier lines, frontline, as well as expand indications beyond AML to high risk MDS would be another group of disorders where there is an even more higher unmet need, if that's possible, than AML.
We are certainly looking at both going late phase as well as expanding our indication. Yeah.
I know you've a lot on your plate now, but there is—y ou alluded to solid tumors and, you know, you've been working on a number of things preclinically. If you can maybe talk about what, you know, is next or could be next coming up the pipe that gets you excited about your platform. Is this something that even could be used in, with in vivo or RT, or just the, what's the breadth of logic gating?
I think building on what Kanya mentioned about collaborations, I think we view this as a technology that's been validated now, and that could plug into other modalities. We have some internal programs under development using T cells for solid tumor applications. There's been some interest in how do you make in vivo CAR better with this, going beyond just maybe the low hanging fruit applications there, going after novel targets with a new delivery modality, or even we have some non-logic gated technologies around like arming cells more effectively to enable expansion, which might be really relevant for in vivo. Hopefully some of those will come out later this year that we can talk about.
Yeah, absolutely. Yeah, I think that's, you know, got through most of the questions here. Oh, we have a couple in the audience.
Yeah.
Yeah.
VSIG2, you have a little neutropenia and you know, you have a little bit of, I guess infections related. Is the inhibitory receptor, you know, partially protective in the bone marrow? I mean, if you didn't do that, you'd get profound neutropenia and profound neutropenic infections. Is that the view? I'm just curious mathematically, how protective is this inhibitory receptor?
Yeah.
I'm sure you've looked at a few of these.
Yeah. Just to clarify, our protective inhibitor for SENTI-202 is endomucin. VSIG2, did you say VSIG2?
Sure.
That's for our other product. Anyway, conceptually—
For VSIG2. Okay.
Yeah. It's endomucin, EMCN. VSIG2 is the protective one for epithelial cells that we have in our, you know, as Tim was mentioning, for some of our future solid tumor programs. Endomucin is, just to give you one number mathematically, preclinically we've seen between 75%-90% protection. For how that translates to clinical is another question. But that's a number. What we have seen is if—w hen you, when you lymphodeplete patients, you know, counts fall, everything falls basically, right? That nadir obviously is gets reported as an adverse event. That's some of the neutropenia thrombocytopenia that we have seen. If you look at our, the—i f you plot our normal cells over time, by day 14 it comes right back. It's like a U.
The nadir gets reported as an event, but it recovers very quickly.
Got it. Then if I could ask one last question. The NK experience, there's been a lot of NK companies—
Yeah?
T hat have come through for AML. There was a Washington University experience. A lot of the CRs that we've seen in the past were in the anywhere from 20s to 40s to 50s. You have in the 40s, it's been always confusing for us on the, on the investment side to know whether that's enriched based on patient characteristics, you know, chromosome aberrations, or is 40 sort of best in class, and this is what we should get used to when—j ust hopefully you could put the NK experience in context versus your drug.
Yeah, happy to. You know, nearly 1,000 patients have received NK-based therapies, right? Almost all of them, single center academic fresh NK cells with no engineering, no CAR, none of that, right? They've seen responses in the 20s, 30s, and they've not been durable. One of the reasons for that is, just if I could get a bit deeper into the NK biology, there are these stress ligands, which cancer cells express. The leukemic stem cells specifically down-regulate those to escape NK surveillance. There's a lot of work that's done in Europe, which shows that very, very elegantly. I think the big difference between what we are doing and the all the other NK experiences, there have been a few other NK CARs as well, is our targets. We are using validated targets.
CD33 is a validated AML target. FLT3 is found on leukemic stem cells. The fact that we have a bivalent target, both of which are validated for AML, I think is leading to what we are seeing in our product, and that's the big differentiator for what we are doing compared to— I'm not aware of one other NK company which has done an AML target.
Got it. Thank you.
Yeah.
Any others? All right. Thank you all for joining us today for the fireside chat.
Thanks a lot.