Precision BioSciences, Inc. (DTIL) Earnings Call Transcript & Summary

Thanks everyone for joining us today on the last day of 44th Annual Goldman Sachs Healthcare Conference. I'm Andrea Tan, one of the biotech analysts here, and I'm super pleased to have the team from Precision BioSciences, Alex Kelly, CFO; Cassie Gorsuch, VP, Gene Therapy. Thank you both for joining us.

Welcome.

Alex, maybe I'll start with you. You recently had a data update for your allo CAR T program. Walk us through -- and it was extensive, but walk us through here what you presented.

Okay. Great. Thanks, Andrea. So yes, last on May 31, we had a full update on our CAR-T -- allogeneic CAR T programs. First, we provided an update on azer-cell, which is our lead program in Phase I/II trial for patients with relapsed -- who've relapsed after auto CAR T treatment. So in the data that we've shown so far, we've seen in that patient population, we saw an 81% overall response rate. We saw a 60% [ CR ] rate. And we saw a duration of response of greater than 6 months in about 55% of patients. So it looks really encouraging, especially in that patient population, which right now, there are no approved treatments for patients who have relapsed after auto CAR T. And if you look at the consortium data that's been published, you can see that those patients have about a 1.8-month PFS. In that population, only 20% to 30% get to a CR. And the mortality rate -- I mean the life expectancy is about 4 to 6 months. So there's a great need in that patient population. And we think with the profile that we've seen with azer-cel that we can provide an effective option for those patients.

Maybe in the context of those statistics that you just provided, what is the target profile or target product profile for azer-cel?

Yes. I think given that poor prognosis with the patients who've relapsed after auto CAR T, I think getting to a 50% CR rate or better and -- I'm sorry, a 50% overall response rate and getting about half the patients to a 3-month duration of response will be a meaningful improvement versus what is available today.

And of the -- maybe the results that you just highlighted for us, that's across all the patients that have been treated in the relapsed population, but maybe dig more specifically into the latest cohort because that is ultimately the dose, the lymphodepletion regimen that you're planning on taking forward.

Right. Yes. So I think when you look at that -- the profile in the last patients that we treated, so we announced that we had 7 patients treated at a dose of 500 million cells with a reduced dose lymphodepletion regimen. Five of those patients were treated at the lymphodepletion regimen that we plan to take forward. That's a dose of flu -- 3 days of flu with 750 milligrams of Cytoxan for 3 days. So that is what we think would be to go forward lymphodepletion and the go-forward dosing regimen in the trial. And in this population, we're above that threshold, we think, is meaningful for patients. We had a 60% CR rate and 80% ORR in the patients that were treated in the [ latest ] cohort. We also really -- most -- very importantly, improve the safety profile. In the previous cohorts, we had seen that there was a high degree of either infections or fludarabine-related neurotoxicity that resulted in some Grade 5 events. So we've completed with our manufacturing process, improvements as well as with the reducing the dose of lymphodepletion, we've seen 0 Grade 3 events or higher in terms of the events that are meaningful for CAR T patients. So we think we've cleaned up the safety profile, and we're still delivering a very meaningful efficacy and response rates. So we're a little bit too early to tell on durability, but we'll keep watching these patients. We think we have a good signal to go forward with.

Remind us what was the median follow-up for these patients?

So the median follow-up for these patients was for the last cohort was less than 3 months. We think that 3-month mark is important. So we'll continue to follow them. But in the meantime, we've requested and been granted a meeting with the FDA.

And there was some discussion around MRD status and you are following that in some of these evaluable patients. Help us or maybe frame for us the use of MRD status as a predictive measure of durability.

Yes. I don't think it's been used widely in non-Hodgkin's lymphoma patients, but we saw a publication from the Journal of Clinical Oncology in June last year -- sorry, June 2021, and that was looking at the ZUMA-1 experience, and it found a very strong correlation between circulating tumor DNA clearance and PFS. So in particular, those patients who were MRD negative were more closely associated with [ PFS ] than any metric that they were looking at.

And do you believe that, that's comparable as you think about the different patient populations that were studied in ZUMA-1 versus the patients that you have here?

Yes. I think it's a fair question, right? We don't have an autologous CAR T product. We have host immune rejection is a bigger issue for us to work through in these patients. But so far, when we look at our experience, our cumulative experience with azer-cel and the 12 patients that we've treated and disclosed last year, we have much more follow-up in those patients. We saw 7 out of 8 of the patients were MRD negative. And we saw very good response rates. We saw a high rate of CRs and also pretty durable responses as well. And in that group, 55% of the patients are getting out past 6 months in terms of duration of response.

And there's been a lot of discussion around 3 months CR rate and that being maybe the metric that you should look for to understand durability. We'll wait and see on that. But maybe when can we expect the next data disclosure?

Yes. So our next update will be after we have our FDA meeting, and that meeting is this month. So I don't expect to ramp a lot more data to share at that point in time. That will be our next step then in the program and give an update on the FDA feedback. But we'll probably have another update on the program. It would be more likely to occur later this year.

Got it. And then at that point, we should have a better sense as to the longer follow-up and the potential for this 3 months.

That's correct. Yes.

You touched on the FDA meeting so we can go there. Maybe provide some background on this meeting, what you're looking to get out of the meeting and what next steps could potentially be?

Yes. So this meeting is our clinical meeting with the FDA. You may recall that in December last year, we had our CMC meeting with the FDA. So we reviewed our CMC plans with the FDA, got a good alignment from them on those plans and so that step has kind of been checked. And the next step, which we're having this month, is the clinical meeting, where we discuss our clinical [indiscernible], what's the size of the trial, what's the length of the trial, will there be a comparator or no comparator and just really iron out and get alignment on the design of the trial. So that will occur this month, and then we'll kind of reflect on the feedback from them, wait for written responses, and then we'll have an update likely in July on where we're going with the CAR T program, especially the azer-cel.

Do you have a sense or maybe early thoughts on what a trial design could look like?

Yes. Look, I don't want to put words in the FDA's mouth. It doesn't usually pay out very well. But I think our proposal will be that we look at a single-arm trial. We think that probably in a range of 75 to 100 patients would be sufficient for that single-arm trial. And the single-arm trial, we think, is relevant here because there is no approved standard of care. There are no approved treatments for the post-auto CAR T patient population. And what they have available to them now doesn't give them very good outcomes. So we think this single-arm trial would be meaningful, and if you reflect back on ZUMA-1 single-arm trial for [ third line ] setting for CAR T. So this is moving to even [ later line ] settings we think is appropriate.

And do you have an understanding -- and this will come from the FDA meeting, but the number of patients or the extent of follow-up that might be necessary to really support moving forward to a registrational trial?

Yes. Good question. I think we'll wait for their feedback on that. We think we have the data package right now between the cumulative experience on azer-cel up to this point, the final dose and the lymphodepletion regimen getting that ironed out, improving the safety profile and maintaining efficacy. We think that we have a package that should merit good discussion with the FDA, and then we'll take their feedback from there.

Are there precedents out there that the FDA has looked at the totality of the data. And as you mentioned, you have many more than just these 7 patients that were in the latest cohort. But are there precedents where the FDA has looked at that versus saying, no, we only want to look at this very specific set, this latest cohort. And in that cohort, you need x number of patients?

Yes. Not to my knowledge, but I haven't been party to those of the discussions with the company. So -- but in our preparation, we're definitely looking at other benchmarks to try to understand where the FDA might be coming from in those meetings in the meeting that we have with them. And I think the team is well prepared to kind of defend the position that we have.

How do you think about -- when you think about that totality of data there, just given that there are the majority of patients that are at different doses from lymphodepletion regimens, which clearly we've seen then impacts safety and tolerability. How do you think about maybe the interpretation of that data set and being able to really read through from those patients to the profile that you'll ultimately see with respect to durability.

Yes, it's a good question. I think the -- where we see the profile of the patients, I think we've got a very good selection so far in the experience of patients with CAR T relapse. I think that we're looking more specifically at the DLBCL patient population. So we're not trying to take folliculars and things like that into this trial. We really want to look at patients that don't have good opportunities for treatment beyond that. So I think that, that will be pretty predictive of the patients we see going forward. And you've been following this program for quite a while. You know that in the first cohort of data that we really shared, we were seeing patients that could be up to 11 or 12 line of treatment by the time they got to azer-cel. We've tried to rein that back a little bit and get into a more appropriate range as we move forward. And now where we are, I think that we're in a good place. We want patients that probably have more or like up to 5 lines of treatment and hopefully, no more than 1 or 2 lines of treatment since their autologous CAR T. So I think that, that's probably one change that we'll see is that we'll [ fewer ] lines of treatment and perhaps fewer between the autologous CAR T and the treatment.

Maybe characterize for us what the market opportunity looks like in a post-auto CAR-T setting?

Yes. So I think that this market right now is probably as big or bigger, and the autologous CAR T market was when ZUMA had approved and we [indiscernible]. We see that there's probably going -- between 2022 and 2025 probably a fourfold increase in the number of patients who will be eligible. And that's just driven frankly by autologous CAR T becoming more well accepted and also more well accepted in earlier lines of treatment. So you end up with second-line approval, which they have, and so more patients are getting treated in second line. But still, we're still seeing 65% of those patients will ultimately fail on treatment and relapse. So that market is going to continue to grow as they move in that setting. So we think by 2025 it could be 18,000 patients available in the CAR T relapse setting in the [ G8 ] market.

Got it. And -- but that's not exclusive just to the, say, like the fourth line or the last line setting, that's across as auto CAR Ts are moving to the...

That's right. They move into earlier, correct.

And how do you think about maybe the landscape that has been evolving in a post-auto CAR T market? I mean, bispecifics, where does an allo CAR T really fit in relative to these other agents?

Yes. I think that we're now fits in is, number one, having [ availability ], but allo will be kind of a game changer for physicians, something that they can use off-the-shelf instead of waiting for a manufacturing process to take place. I think the other thing is that with an allogeneic CAR-T, the goal was a [indiscernible], right, to get a single treatment that results in a long-term durable response. And I think with bispecifics, I think they're attractive products, but they really are not providing that potentially curative response. They're potentially providing a chronic therapy.

And do we have a sense as to the number of patients, I guess, maybe when they're progressing post-auto CAR T? How many are you still retaining that CD19?

Yes. What we see in our experience, and we screened a lot of patients, is that 85% of patients still have CD19 positive disease after they've relapsed from CAR T. So we think that, that's still a very attractive place for azer-cel because [indiscernible] CD19 positive patients. Now that said, we did have one patient in the last cohort who responded to treatment. They had a 60-day CR and at day 90, they had disease progression due to antigen escape. So we do see it, but we think it's the minority of patients, more like 15%, as opposed to majority of patients.

And then as you think about auto CAR T moving to the earlier lines, as you mentioned, what does that imply for your strategy as a company in terms of the trials you would need to run and your capacity interest even keep following them to earlier lines.

Yes. I think for azer-cel, our focus is on the CAR T relapsed patient population. I think when we think about trying to compete with auto in earlier lines of treatment, that's where stealth cell or PBCAR19B we think will really play a role.

Maybe another way to ask that, as auto CAR T goes, it's in second line now, if it goes to the [ front line ], would you be interested in running studies for azer-cel in that third-line setting, second-line setting and kind of moving forward?

Yes, I think we could look at that. But I think we also have stealth cell, which might be a better profile for that earlier on setting and the ability to delay the host immune response might be more evident as healthier patients.

Maybe let's jump to 19B then, which you've been referencing. Remind us how it's different than azer-cel?

Yes. So PBCAR19B, we call it stealth cell because it has the same basic construct as azer-cel, where we're knocking the CAR receptor into the track locus. The -- but it has 2 other features as well. Number one, it has about a 90% knockdown beta-2 microglobulin, which we think will help offset T cell rejection, but also by only doing a 90% knockdown instead of a complete knockout, we think that it will also help prevent NK cell recognition and rejection. So that's one feature, the beta-2 microglobulin knock down. The second feature that we've added to stealth cell is we've added an HLA-E transgene, and that is designed to offset improvement NK cell reject. So I think that that's what we intended with the program based on the way we've designed it. And so far, the data looks like it's meeting that proof of concept of being able to delay host rejection.

How does this compare? Maybe just the early profile you've been seeing. How does that compare to maybe other immune-cloaking approaches?

Yes. I think other approaches that we've read about are using a complete knockout of B2M, beta-2 microglobulin. And we think that, that is an approach that might lead to recognition by the NK cells that this is a foreign cell, this is not the host, and it might lead to more and faster NK cell rejection. We've seen that in our laboratory experiments and in the early data with stealth cell, we do see a delay in the NK cells ticking up as well as the T cell.

And then in terms of the data that you did present for 19B very early, a few patients, but walk us through what you saw there?

Yes. So we saw that proof of concept coming through in terms of the ability to delay host immune rejection. So that was rewarding for us. And remember, this is at a dose of 540 million cells with the same lymphodepletion regimen that we're studying with azer-cel, the cy750 regimen. So in that group, when we look at the DLBCL patients, we see a really high response rate, and we also see a high CR rate in those patients as well. And I think that we see data that looks like it could compete with autologous CAR T in terms of efficacy, CR rates. And hopefully, one time, we'll see durability as well. We're too early right now to make the call on durability. But we are seeing a high degree of MRD negative CRs, which we think is a leading-edge -- leading indicator for what we'll see in terms of durability.

Same question for as I had for azer-cel maybe when can we expect the next update here?

So I think the next update on stealth cell will probably be more likely to be later this year. So it probably won't be subject to the update we gave on the azer-cel FDA meeting, but we'll let this data continue to mature and continue to follow these patients. So I think that will be an important parameter. I think with stealth cell and azer-cel, the time line probably matters more. And I think for azer-cel because these patients don't have other options, a shorter duration of response is probably really acceptable. I think for stealth cell because we want to displace autologous CAR T a longer-term duration of response is much more appropriate. So we'll be looking at 6 months as a key indicator, but I think one of the places that allogeneic CAR Ts have fallen off in the past is that even when they get somebody to 6 months of duration of response, it tends to fall off after 6 months. And I think the experience with autologous CAR T tends to show that what you get at 6 months tends to be retained at 12 months. So it's a good predictor of 12-month response. I think with allo, we need to still prove that out.

Do you have a sense maybe in that context, if 6-month CR is not sufficient? Maybe what is the length of time that would make you feel comfortable that each really do have a durable construct?

So I think 6 is a key benchmark just to see if you're above that 35% number at 6 months, but then we'll be watching after that to see how long is it durable. And if we can see the same kind of duration and those patients persist at 6 months and also at 12 months, I think that we'll be in a good place to know that, that 6-month mark with allogeneic CAR T is still a good predictor of the long-term durable response.

And as you think about bringing forth azer-cel as well as PBCAR19B, there's been some talk around exploring a partnership or someone to take on 19B, maybe provide a little bit more color around that and what you would specifically be looking for in terms of a partner or a collaborator?

Okay. Yes. So I think partnership would be really helpful for us as it relates to stealth cell and 19B. Remember, the goal is to displace autologous CAR T. And to do so, we think it's going to be a head-to-head study. We think it will be a large study, and it will also be a time-consuming study. And unfortunately, it'll also be an expensive study. So I think with those 3 things, it's probably not in Precision's best interest to tackle that alone. We think ultimately commercializing in the -- against an auto CAR T., we'll probably take a reasonably sized sales force and commercial effort as well. And that's something that may be better off in the hands of a bigger partner. So I think when we look at our overall partnering strategy, whether it's in vivo or ex vivo, the approach is like, is there a partner that can enhance our capabilities, either commercially or doing bigger clinical trials? Or does it enhance our [ physical ] abilities -- capabilities. And I think that for stealth cell, a partnership would definitely help in both of those [ drugs ].

Perfect. Well, maybe we can jump over to the in vivo gene editing, which a lot of focus, obviously, is on the allo programs, but you guys also do have this in vivo portfolio. Cassie, let me bring you in here. Maybe speak to the efforts across this portfolio. Notably, you have both partnered programs, you have your own in-house programs, maybe describe for us first the overview and the setup of that portfolio.

Sure. Yes. So as you mentioned, we're really a gene editing platform company. And so at our core, we're a gene editing company. We apply it for allogeneic CAR Ts, and we apply it for in vivo gene editing as well. On the in vivo gene editing side of the house, our lead program is our internally-owned HBV program. We have a number of partner programs as well. Sort of first in line is our OTC program, which is partnered with iECURE, and maybe we'll jump into that a little bit more later. Additionally, we have a partnership with Eli Lilly, and we recently disclosed some really exciting data for one of those programs. It's our Duchenne muscular dystrophy program. So that's the only disclosed target at this time with Lilly, but of course, we have 2 other programs that are ongoing with them. And then we have a really exciting partnership with Novartis that is really, again, sort of taking advantage of some of the unique properties of our gene editing platform. And so as we think about sort of all of those together, it's busy times. There's a lot going on at Precision, but I think there's a lot of really exciting data coming out, and we're learning a lot about how to best utilize this technology that we've built for what it's designed to do best.

Maybe let's start with the DMD data since you did just present it at a medical meeting. Walk us through the data that you showed and how meaningful is it in the context of maybe other assets that are being developed in this space?

Yes. We recently shared some proof-of-concept data for DMD program at ASGCT in May. I think the highlights of the data are really -- this is -- we're excising -- the therapeutic approach is that we're using 2 ARCUS nucleuses to excise a hotspot region of the dystrophin gene. And we're excited about this therapeutic approach because 50% of patients have mutations that lie within that hotspot region. And so this therapeutic approach can apply broadly within the patient population compared to others in the space that are really aimed at single exon skipping approaches, which will apply to a much smaller population. But the data that we shared showed that excising that region, we could do that with very high efficiency in a DMD mouse model and that excision led to a functional response that was 86% of wild-type control animals, which we think is a really exciting improvement in terms of muscle force output. The other piece of data that I think really is a differentiator for us is that we demonstrated that we can edit in Pax7 muscle satellite stem cells. And these are the cells that give rise to new muscles. And so when you think about durability, those are the cells that are most important to effect. And this is something that others in the space have really struggled with. I don't think there's very good evidence that the microdystrophins, for example, actually affect these cell types. So we think that piece of data is really important and a key differentiator for our approach, and we'll speak to long-term durability.

And when you think about excising this hotspot that you've spoken about, how crucial is it that you have the ARCUS nucleus that's doing that cutting versus maybe another editing tool?

Yes, great question. We think this therapeutic approach really takes advantage of 2 very unique properties of the ARCUS platform. First is the size of an ARCUS nuclease. They're extremely small compared to other editors in the space. They're so small that we can actually package both nucleases that are required for this excision in a single AAV. That's really important, especially in the DMD space. We know that to reach the tissues that are most affected in DMD patients that a lot of what we see is high dose AAV in the clinical setting. And there's obviously some pretty high risk associated with high-dose AAV. So being able to package all of the components that we need to complete that edit in one AAV, we think is essential to make sure that we have a good safety profile. The other piece that I think is important for this excision in terms of using ARCUS is that the type of cut we create is a 4-base pair of 3 prime overhang, and that overhang allows the gene to come back together perfectly. If you create a blunt cut and you see this in other's data that have presented a similar type of editing strategy, the way that those cuts come back together is a little bit more stochastic. You can get other types of genomic rearrangements. For us, that perfect [ religation ] is the predominant outcome. And we think again that, that is a really important safety feature.

And maybe one last question on this program. Just where do you stand in terms of bringing a candidate forward towards the clinic?

We're in our preclinical phase right now. The data that we shared was proof-of-concept with early generation nucleuses for this program. This is a partnered program with Lilly. So we, along with our partners, are really excited about this data and continuing to progress the program. But at this time, it's a little too early to say how close we are in terms of [ continued ] clinical development.

Maybe switching over to your partnership with Novartis. The decision to explore sickle cell, beta thal in the context of a field that is rapidly expanding, it's evolving, there's products that could be approved quite soon. How did that decision come about in terms of how Precision and Novartis decide to go forth in this program?

The Novartis approached us looking for a partner for an in vivo gene editing approach for sickle cell beta thalassemia. And they came to us, again, I think, largely because of some of the unique properties of ARCUS. I think, again, if we think about in vivo gene editing and getting to those hematopoietic stem cells, delivery is essential. Delivery is a challenge there. And so you want a tool that is compact, so you have fewer things that you need to deliver to the cell types of interest. So ARCUS again, is small, it fits nicely in an AAV vector, have a lot of different options for delivery. This particular program is aimed at inserting an anti-sickling gene, and so that's sort of the other side, I think, of what makes Arcus attractive for this approach is our high-efficiency gene insertion. Novartis was previously partnered with Intellia for an ex vivo approach. I think as the field for that therapeutic area has really matured, you can see the attractiveness for an in vivo approach, an off-the-shelf onetime administration that's ready to go, similar to how we think about our allogeneic CAR T cells. And so when it came to an in vivo approach for this indication, I think ARCUS was a really attractive solution for that problem.

And maybe continuing with ARCUS and maybe the benefits or the advantages to this particular nuclease, how do you think about the risk of off-target editing? And how is maybe -- maybe speak to that?

Yes. So specificity is absolutely critical for all in vivo gene editing. It's really important that you understand the specificity of your therapeutic approach for ensuring safety in patients. So we take it very seriously. I think an important feature that we don't talk about a lot is that ARCUS is a protein that has a protein recognition motif for DNA. And so the way in which it interacts with DNA is fundamentally different than a CRISPR-based approach, for example. And because of that protein DNA interface, that allows us to utilize protein engineering to actually optimize for specificity. And that's what we do in our early stage. Development is as we're looking at new targets, we're looking at how do we ensure that our protein engineering efforts are aimed at increased activity, but also really honing in on specificity. On the flip side, if you're using like a guide RNA, for example, there's not as many optimization levers when thinking about specificity and how to improve specificity other than moving where your guide is. And so we think that protein DNA interface is an essential and really attractive feature of ARCUS for actually ensuring we can actually do something about improvement on specificity.

And then as you think about maybe where the field is going, there's -- now there's base editing, there's prime editing, there's gene writing, there's epigenomic editing, how does ARCUS fit into that entire toolbox.

Yes. It's very exciting to see all of the growth in the gene editing field. And as you mentioned, there's new technologies popping up all over the place. I think ARCUS has been -- ARCUS is based on homing endonucleases, which were really the first gene editors in the space. Our cofounders were some of the first people who actually were able to achieve gene editing using what was the early version of ARCUS. I think since then, we've made a lot of improvements in the platform. We've really learned a lot about how to utilize this technology. And I think the thing that maybe sets it apart a little bit from some of the other types of editing that we're -- editing technologies we're seeing come out in the space is that those are really geared at a specific type of edit. They're not necessarily as versatile. ARCUS is able to do a lot of the same things that those other gene editors can do. So base pair changes, small DNA insertions, large DNA insertions, like gene writers, but ARCUS -- the platform can actually achieve all of that using the same fundamental technology. I think it has some advantages in terms of the flexibility for the type of edit that you want to create.

And then you've got a number of other programs that we haven't fully touched on, maybe highlights from those where you stand with them and a preview, if you're able to, into the R&D Day that's coming next month.

Yes. We're really excited about this R&D Day that we're planning for July. Our goal and that is really to provide a really in-depth overview of what's going on in the in vivo side of the house. And I think you can expect to see updates across our named programs as well as potentially some new things, new data I think that's probably about as much as I'm allowed to say right now, but...

You've got a long...

Yes, I think some new data, potentially some new programs. It will be an exciting event.

Any highlights from the OTC, HBV, PH1 programs that you'd like to share?

Yes. So our OTC program is a partner program with our iECURE. It's -- that program is really in their hands, driving towards that IND/CTA filing that we expect later this year. So that will be an exciting time for us as well as proof-of-concept first time for in vivo gene editing with ARCUS. So we're excited to see iECURE continue to progress that program. On the HBV front, we've got some big goals this year associated with the program. We are planning on nominating clinical candidate later this year, hoping to share some of the supporting data towards that candidate nomination hopefully towards the end of the year as well. And then on the PH1 front, that's a program that we've decided to deprioritize at the moment. We're really focused on some of the -- a data-driven decision to really prioritize some of our other programs, like HBV, where we're seeing a lot of really promising data come out.

Alex, maybe a question for you here. Obviously, a lot going on across both parts of the house. How do you think about resource allocation? And are there opportunities, say, with PH1 where it's been deprioritized here? Is that a prime opportunity to then look for a partner or someone to take that?

Sure. I mean I think that there are partnering discussions that are going on kind of on a continual basis. Every time we put out new data, for example, on the CAR T program, we get some new inbound interest in CAR T, which is good. But in vivo has a real steady drumbeat, whether it's going to the bio meetings, whether it's going to ASGCT, whether it's going to other conferences, we see a steady interest in the in vivo programs. And I think that a lot of interesting conversations have taken place, some are still ongoing. Nothing really to flag right now. But I think that there's a keen interest in what ARCUS can do and what's different. Cassie mentioned the Lilly agreement, the Novartis agreement. One of the things that Novartis really wanted, as Cassie said, was insertion, being able to do gene insertion in vivo. And it's something that ARCUS probably is better suited for than maybe any other gene editing tool just based on the cut that Cassie talked about and how it comes back together. I think that we've got a lot of opportunities in many disease states that we could treat with ARCUS.

Can you remind us on your cash position as of the end of 1Q and maybe what that runway is expected to support?

Sure. We had $158 million in cash as of the end of the first quarter. That's enough cash to carry us through the first quarter of 2025 and fund all the programs that Cassie's talking about on the in vivo side of the house, obviously, for stealth, if we want to keep advancing that, we're going to need to find a partner there. But I think that we're in good shape. But we continue to look for ways to you'd be very physically disciplined, reallocate resources. You've seen that historically in terms of -- we've made some tough decisions about some programs, and we decided not to continue pursuing. And so now we're very focused on 2 CAR T programs, and we're focused on HBV right now, and then we'll have some other things that we want to talk about at the R&D Day.

Perfect. Well, looking forward to it. Thank you both for joining us. Thank you, everyone.

Thank you, Andrea. Thanks, everyone.