Beam Therapeutics Inc. (BEAM) Earnings Call Transcript & Summary

All right. Thanks again for joining us at our health care conference. I'm Eric Joseph, senior biotech analyst at JPMorgan. And our next presenting company is Beam Therapeutics. And it's my pleasure to welcome John Evans, CEO, to tell us a little bit about the company. Before I hand it over for the presentation, I just want to note that there is a Q&A after the presentation. [Operator Instructions] With that, John, thanks again for joining us today.

All right. Thank you, Eric. Hello, everybody. Great to be here today. Sorry, not to be in-person San Francisco, but nonetheless excited to be entering the new year, 2021. Okay. So I'm going to walk you through a little bit about Beam today. I'll be referring to the slides, which I know you have access to as well. And Beam is really trying to create a new class of precision genetic medicines using base editing, and I'm going to tell you about that technology today. I will be making forward-looking statements today. So on Slide 3, our vision is really to provide lifelong cures for patients suffering from serious diseases. And we really believe that we are on the cusp now as an industry of a new era of onetime curative therapies. This, of course, begins with gene therapy. Now it's moving into gene editing and within gene editing, base editing is rapidly emerging. This will include gene editing for both rare and common disorders. And ultimately, as we have seen with the mRNA vaccines, we are building a platform to create rapidly programmable precision medicines. Once we get this engine set up, we can very quickly print out new programs with very rapid time lines to exploit new biology. So on Slide 4, really, the gene editing field to date has been characterized by what we call nucleases. And these are the scissors of the genome. So this includes TALENs and zinc fingers, but also more recently infamously, CRISPR. And what these tools can do, which is amazing is you can target a single address within the genome out of 3 billion bases. But once you get there, you only can do one thing is just make a cut. And so that blend double-stranded break occurs and really it's up to the cell to put the pieces back together at the end of that point. So the targeting is really good, but the ability to control what you do to the gene is relatively limited. So with base editing, what we want to do is something that's more precise. We want to be the pencil of the genome, where we can race any letter we want, write in the letter in its place, and you didn't disrupt the sequence around it. So the targeting, the precision, the control and the efficiency of what we're doing is going to be much higher. And that's exactly what we can do with this base editing system. So on Slide 5, we can show the construct itself. So this does use CRISPR for targeting, which we think of as CRISPR's best feature. So shown here in gray is the CRISPR protein, and that is exactly the same as many CRISPR applications. We target it with a guide RNA. And that guide RNA includes an address of 20 bases in the genome where we're going to search and then bind. But now the differences appears. We have changed the CRISPR, so that it no longer creates a double-stranded break. It merely binds that recognition site and opens the DNA out. And we have tethered to it a second domain called the deaminase, and that's shown here in blue-green. And these deaminases that we use can only edit single-stranded DNA. They will leave alone all of your double-stranded DNA. But once they're presented with a short segment of DNA as the CRISPR binds, they'll edit very efficiently. And if we use an adenosine deaminase, we will edit As, we use a cytidine deaminase, we edit Cs, those are the A-based editor and the C-based editor that we have in our portfolio. So the advantages of the system are quite numerous. So first, we're directly editing disease-causing genes within the coding region. The applications are very versatile. Of course, gene correction is of interest of point mutations, where you have a single letter misspelling in the gene, are by far the most common class in mutation causing genetic diseases. So things like sickle cell anemia, Alpha-1 antitrypsin deficiency, we would aim to directly fix that gene by just changing the one letter that's wrong. But equally, we can do a lot of other things with this tool as well. We can activate any gene by targeting regulatory regions, we can silence any gene in a variety of ways with a single letter change, all without cutting. We can also do multiplex editing where we can edit simultaneously across multiple positions in the genome. And without cutting, we don't worry about the chromosomes being put back together again in the wrong order, we can push the boundary of multiplex editing quite significantly. Specificity is really attractive. We have load undetectable off-target profile with this system. And just in general, it is a gentler -- the cell almost doesn't know it's been edited because it can't detect the editing event, primarily that's because we avoid a double-stranded break that's inherent in the nucleus editing. Finally, efficiency, this is very efficient. That's partially because it's chemistry. Once we're in the cell, we will bind to the DNA, we'll open it up, and then this enzymatic reaction is a natural and very quick one that we can even tune and optimize. And so we get very high levels of editing efficiency, anywhere from 70% to 80%, 90%, often even more efficient than what's possible with the nucleus. So ultimately, a very differentiated proprietary next-generation gene editing technology in base editing. So Slide 6 shows you how we have adopted a strategy to really bring this forward ambitiously across all of the different delivery modalities that are currently available in genetic medicine. So that includes electroporation of cells outside of the body. So this would be blood cells in hematology, and that will bring us into sickle cell disease. T cells in CAR-T to treat cancer. And here, we're doing initially some leukemias. Nonviral vectors, which will go in vivo, now to the liver initially, where we're treating Alpha-1 antitrypsin deficiency and glycogen storage disorder, now precisely correcting the point mutations in those disorders. And finally, viral vectors like AAV, which can bring us to the eye, the CNS and many other places, initially targeting Stargardt disease for precisely correcting the most prevalent permutation there. All of these programs are moving forward in parallel and moving quite quickly. And significantly, we're the only company that has gone this broad in delivery. So as each of these delivery modalities comes online, we now have the ability to double down in each of these tissues so that next-generation programs can quickly advance with much higher probability to actual success once these initial programs in each area have succeeded. Obviously, ex vivo is moving a little faster because there's no vector engineer, as is true with the in vivo portfolio. But really all of them moving now in parallel and making quite rapid progress. In terms of milestones on Slide 7, 2020 was an amazing year for Beam. We achieved, I think, more than I expected, honestly, in terms of our ability to move forward. We have 3 development candidates now named: our ex vivo portfolio, BEAM-101, for the up-regulation of hemoglobin and sickle, that is also now in IND-enabling studies. BEAM-102, really the first industry programs to directly correct the sickle mutation itself, over 80%. A really unprecedented result, also a development candidate and moving forward just behind 101. And finally, BEAM-201, a CAR-T program with 4 different edits to provide additional functionality in that program as well as in vivo proof of concept that we published late last year. So all of these programs now moving forward, quite quickly in parallel. BEAM-101, we continue to be confident is an IND filing in the second half of this year. And so that's moving forward quite quickly. We've been to the FDA. We've gotten good feedback from them, and it's quite clear, which way we need to go. BEAM-102 and 201, moving along just behind that and not far behind. So those will be next into IND-enabling studies, and we'll be able to give additional feedback on IND filing time lines as we get it. For the liver portfolio in vivo editing, we've shown in vivo proof of concept of direct correction with Alpha-1 and GSD. Next steps for that will be show LNP data this year. We should expect early in the year of primate editing, which we're quite excited about. That would set us up to deliver our first liver development candidate on the liver portfolio later this year. Finally, the AAV program is moving forward, and that will also be into primate studies earlier this year in collaboration with the Institute of Ophthalmology in Boston. So before I dive into the portfolio, let me just step back a little bit and tell you about where we're going with Beam. Ultimately, we are establishing what we think of as the leading platform for precision genetic medicine. And we're so excited about this vision. So of course, base editing is a big part of what we do, highly differentiated technology. It will form the dominant part of our portfolio for the long term. But that said, we have other editing systems as well. So we also have a nuclease family called Cas12b. We can do RNA editing. We have access to a new system called prime editing. So already, we have one of the broadest suites of editing technologies in the industry. That said, none of this matters unless we can deliver these technologies. And so here again, we have one of the broadest suites of delivery technology already. We can do autologous cell editing as in sickle. We can do allogeneic cell editing as in CAR-T. we can do mRNA at a very pure level and a very high sophisticated level. And of course, we can do nonviral and viral vectors. Finally, none of this matters unless you can manufacture it. So we announced last year, we're going to be building our own 100,000 square foot manufacturing facility. This is going to be in North Carolina. It'll be opened by 2023. This is, again, the largest manufacturing capability that we believe is available in the editing space. When you put all this together, and it will establish in Beam really the full stack of capabilities that you need to create any precision genetic medicine you want for any biological situation. And we're very excited about this. It will allow us to go very deep in the franchise areas that we want to win in such as sickle cell, Alpha-1 and others. It allows us to break new ground and new tissues using these delivery technologies, where, again, we can identify high-value programs that we can develop ourselves. But also makes us a one-stop shop for partnering either with other innovator companies who may have expertise in biology or disease areas or with larger pharmas who want to establish a footprint in this space. So we think the power of bringing all of this together under one roof is quite significant. We've already made incredible progress on the editing side, and we do expect to increase our investment on the delivery side overtime over the course of the next year. All right. So let me dive into some of the programs now. So on Slide 9, so our ex vivo delivery. Here, we're taking cells out of the body, electroporating them and putting in the editing machine room. So this includes the guide RNA to target within the genome and then the editor RNA, which will encode for the editor. All of our programs are using mRNA. We think it's a really attractive, synthetic and scalable way to deliver our editors into these cells. So on Slide 10, in sickle, we're going to be doing what we call autologous cell processes. So here, we are taking disease cells out of the body, we're editing them using this electroporation technology. We then ablate or condition the patient to get rid of the other old cells and then we reinfuse these corrected cells for long-term engraftment to really repopulate the blood system. So in sickle cell disease, Slide 11. Of course, this is the most famous point mutation in all of genetics. A single T is misspelled as an A, and this causes a new form of the globin called sickle globin to occur in the cell, which then polymerizes, deforms the cell and creates these painful crises and organ damage, and ultimately, early mortality. Very significant number of people suffer from this disease. So we have 2 best-in-class approaches for editing sickle cell, BEAM-101 is the first. So here, we're taking advantage of human genetics to create point mutations in the regulatory regions of the fetal hemoglobin genes. Now fetal hemoglobin regulation has been practiced by others. It has clinical proof-of-concept now. It's clearly going to work and be transformative. In our case, we can do it much more directly than others because we aren't cutting. We can go directly to the on/off switch of these genes and put in these point mutations that turn them back on. In this case, these same mutations are actually found in people who have persistently high levels of fetal hemoglobin throughout their life. And if they also happen to have simple mutations or beta-thal mutations, they are protected from those diseases. So we're levering that clinical genetics for the strategy we take here. So it turns out that using a base editor to install these point mutations is quite efficient. So on Slide 13, we show you that data. We're well over 80% base editing at these 2 different promoter regions. So by definition, this is a multiplex edit because the same guy will bind to each of the promoters of these 2 duplicated genes. In addition to the high editing, we get a very robust biological response. So we see very high levels of that, over 60% editing. And that is, as far as we know, the highest level of up-regulation [ vector ] that's been reported in field. So the combination of these 2 things, a high number of cells that's been edited and each sell a high level of F is going to give patients the most robust protection from the activity of that sickle globin from causing polymerization because this fetal hemoglobin basically blocks that polymerization. At the same time, by switching so heavily to the fetal form of this globin, we're also simultaneously switching away from that mutated version, as you see here on the right. We're actually lower than 40% levels of HbS. And that is, again, the strongest and lowering of HbS among the fetal hemoglobin approaches that we've seen in field. We have recapitulated these results in vivo, shown on Slide 14. So here, we do editing of human cells, and then we can graft them into banks long term, and we track them overtime to make sure the engraftment has health. So on the left-hand side, you see very strong engraftment, over 90% human timers in up to 16 weeks. See very tight error of ours. It's a very robust and reproducible process. And particularly, you see no difference between edited and unedited cells in the engraftment. And that speaks to that milder edit that we made, which really preserves cell viability. In the middle, you can see the very high level of editing on all of the different lineages within this cell population, including all the ones that contain the stem cells. And so that again speaks to base editor's ability to really uniquely be consistently applicable in any cell state, dividing/nondividing, any type of biology that's active in the cell that the edit will still occur. And finally, of course, recapitulate the high level of S up-regulation here, over 65% have -- with the A base editor. So I'm very excited about BEAM-101, moving now rapidly towards the clinic and be an IND filing later this year. So BEAM-102 on Slide 15 is our second approach. And here, we're doing, of course, what you'd like to do, which is go directly at the sickle mutation cell. So in this case, we don't have a base editor that can go A to T, and revert to the form of hemoglobin that many of us have. But luckily, our A to G editor can still turn out an edit that produces a form that is normal. And it's called Makassar or HbG. About 0.1% of the in-population has this. It is normal. It binds oxygen normally, it does not polymerize or form sickling. So with this editor now for the first time, we are literally creating cells that no longer have the ability to create sickling. They're eliminating the mutation itself. And so you can see that editing data on Slide 16. Over 80% editing of this directs causative point mutation. When we go a layer deeper and look at the allelic burden of these cells, we can see that actually over 70% of the time, we're actually editing both copies. So we're fully removing the sickle mutation from these cells altogether. In the next 20% of the time, we are adding at least one copy. And that would be more of a profile like a patient who had sickle cell traits who also don't get disease. So that means that over 93% of all the cells are potentially cured of sickle. And that is, again, an unprecedented result. When you look functionally what this does on Slide 17, you see what you would expect, which is that as we dial in the edit on the left-hand side, we are, again, one-for-one, replacing a new copy of that gene with the old mutated copy. And so by the time we've gone to 80% editing, we're down around 10% sickle globin. So again, an unprecedented lowering of the positive protein in these cells. And of course, on the right-hand side, you can see the low oxygen conditions. We are indeed eliminating the sickle pheno type altogether. So very excited about this as a potential cure for sickle. I would say, in both cases, not only are we moving forward to the clinic, we'll be doing kind of the standard trial using standard disulfide conditioning to ensure we can establish proof of concept. But we're quite excited as well about the ability to move on to next-generation conditioning regimens and improve the ability to deliver these editors to more people in a nontoxic way. We think the entire field is going to move in that direction, and we are hard at work to make sure that happens because it will pair really well with our best-in-class editing. Okay. Slide 18, let's shift gears now to the CAR-T and T cells. So obviously, CAR-T has been a revolution in cancer care, the ability to take immune cells and retarget them directly towards the tumor itself. But beyond a couple of really notable early successes, we have more than we want to do. So what we'd like to do is make many more edits in these cells and engineer more function into them. So what that will mean is knocking out things and silence these proteins using editing. So now we're silencing and we're silencing in a multiplex fashion. So on Slide 19, what you can see here is when we do on a single electroporation, we put in the RNA for the editor once and then we put in as many guide RNAs as we want to target within the gene. It's a single reaction. We don't do this in serial. It's all one parallel single electroporation. And as many guys we put in, we will then simultaneously edit once these editors are expressed in the cell. So it's a very powerful approach for multiplex editing, putting in a lot of functionality into these cells. And then, of course, we bring them back into patients. In this case, these will be allogeneic products, so we're going to take cells out of a healthy donor and be able to freeze them and then ultimately give them to many patients as soon as they need them off the shelf. So this is Beam's future of cell therapy. So Slide 20. BEAM-201 is our candidate here. This is the first quad-edited cell in the industry. We're making 4 different edits to prevent graft-versus-host-disease to make the cells allogeneic to add efficacy by avoiding immunosuppression. And ultimately, the target here is CD7. In this case, the antigen we're targeting in the tumor is also expressed on our T cell. And so if you don't edit that target out, which is the fourth edit, then you have what's called fracture cell, where the CAR-T cells are killing each other before you have a chance to infuse them and hopefully kill the tumor. So by knocking it out, they will ignore each other, and they will only target the tumor. So this program is moving forward very nicely. And you see here the dramatic efficiency with this system. So 4 edits at the same time, single reaction, each edit is in the 96% to 99% range efficiency for knocking down these proteins. And over 90% of the cells receive all 4 edits. So that is a remarkable result, and the cells are very prudent. You can see on the right-hand side, a very aggressive in vivo tumor xenograft in these T cell leukemias. And you can see a beautiful dose response of beginning to eradicate these tumors. So very excited about what this program can do. Now on Slide 22, you can, in theory, knock things out with nucleases as well. And other companies are doing this, and it can be efficient. The challenge is when you make multiple edits at the same time. So you get 2 issues with them. On the left-hand side, if we make simultaneous triple edits to 3 genes with a nuclease and then you detect for translocations. So this would be the cell, putting the pieces back together in the wrong order, you can readily detect those chromosomal rearrangements using -- when you use [indiscernible] nucleus. With the C-based editor, we find none. That makes sense because we're not making any double-stranded breaks. Furthermore, on the right-hand side, as you add edits going left to right, going from 1 to 2 to 3 edits with a nuclease, you start to see a decrease in viability of these cells. And whereas with the C-based editor shown in blue, we don't see any decrease. And that's again because the cell doesn't really even notice the editing is happening. There's no detection of it, whereas we create double-stranded breaks across the genome or else the genotoxic events key to their pathway turns on, the cells will start to arrest and can even die. Both of these problems get much more acute, the higher you go in the number of edits whereas, again, with base editing, we don't face that issue. So we think that base editing is a really powerful paradigm for cell therapy for the future of highly engineered cell products. Okay. Now shifting gears to our next area, which is in vivo editing. So now on Slide 23, we're going to use lipid nanoparticles for nonviral delivery of base editors. And so here, we are formulating the RNA, putting the editor and the guide RNA into that synthetic lipid nanoparticle, infuse it directly into the patient, it will travel at least to the liver. There may be other places we can go as well over time and edit. This is scalable. And in this case, you can also think about redosing if you ever need to. So our lead indication here is Alpha-1 antitrypsin efficiency. So this is, again, a huge population, a very significant unmet need. And again, as a sickle, every patient has the exact same single-letter misspelling in their gene. They have an A where there should be a G. So now our A-based editor can revert that back to normal with a single A to G change in that coding region of the gene. If we can do that, we can create 2 important benefits. First, we will stop this gene from creating mutant protein that builds up in the liver and creates toxicity and ultimately can lead to liver failure. But at the same time, for every copy we correct, not only is it not creating toxic protein, it is now creating normal protein that can be successfully secreted, which is what the liver is supposed to be doing. And it would then travel to the lines where it will protect the lungs from degradation in patients with this disease because of that loss of degradation or have gradual emphysema, loss of lung function and even end up with double lung transplants. So ultimately, base editing is a potential cure that can address both sides of the equation, the lung phenotype and the liver phenotype, and that is unique in the editing field. So for editing on Slide 25, you can see our editing target. It is this blue A in the gene sequence. And for every time when we can successfully edit that A back to a G, we will create a beneficial effect. In this case, we can occasionally -- a minority of the time, we can occasionally hit a second A as well, given the window of editing that we have in this editor, we call that a BI standard edit. As long as we hit the blue A, even if we hit that second A, we still get a beneficial [ effect ]. It is a functional gene. It does not create toxicity of the liver, and it can successfully secrete, go to the lung and prevent degradation. So you can see here, 45% beneficial wheels are created, and this is in patient fibroblast carrying both copies of this Z mutation from human patients. So again, a remarkable level of editing here. We've taken that in vivo on initial experiment in the disease model of Alpha-1 antitrypsin efficiency shown here on Slide 26. Here, we've got about 16% editing initially after a single delivery of an LNP with Beam formulation of this editor. Over time, those cells appear to have a selective survival advantage, probably because of the lack of toxic protein buildup. And this has been seen before in other settings. And so we actually see those cells grow out as a population within the liver over time and reach about 30% of the cells by the end of this measurement period. So in addition to editing, we can look to them the functional consequences of this correction. And indeed, we do see on the left-hand side of Slide 27, reduction in liver aggregates from the correction and. On the right-hand side, you can see an almost fivefold increase in circulating normal Alpha-1 antitrypsin protein in these mice because that gene has been corrected. And we do believe that this level of correction and upregulation of the normal protein would be sufficient to have a transformative impact on these patients and get them up into a normal range of Alpha-1 antitrypsin. You can see also the power of the biomarker that we will have in some development, where very quickly after editing, we will be able to detect levels of normal protein for the first time in these patients' blood systems. So very excited about this program, highly differentiated and could be a first-in-class approach to a cure for Alpha-1 antitrypsin deficiency. On Slide 28, our second program in the liver is ligand storage disorder, devastating disease where patients can't fast. So they can't reprocess sugar back in -- out of their liver. And so they need to eat basically ever 3 to 4 hours. And that includes overnight. And so you literally can't sleep through the night. You have to wake up constantly and keep your sugar levels high. And if you miss a feeding, then you can literally die at hypoglycemia, it's very sudden. So a terrifying disease. In this case, we can target the 2 most prevalent point mutations in this disorder, and both are directly correctable, back to normal with the A to G editor. So on Slide 29, you can see in vivo results, again, showing 70% correction and 40% pressure, respectively, of these 2 different mutations. And clearly, that will be in excess of the therapeutic threshold for cure, which we believe is around 11%. Unlike gene therapy, here, we don't worry about this wearing off over time. This would be a true durable edit. It would be endogenously regulated because we're fixing the gene and its normal position in the genome. And significantly, we can also treat very early in life because as the patient grows, their cells will divide, they'll carry the edit with them. Whereas with the gene therapy like AAV, that signal will be diluted out. So you can't treat infants or very young children. So a significant number of advantages for this approach over gene therapy and a potential cure for patients who need more options. So finally, let me wrap up with our viral portfolio. So here, we're delivering with AAV vectors. So AAVs have a packaging capacity limit. And our base editors, which are a little larger than CRISPR, are going to be a little too big for one AAV alone. But fortunately, what we can do is we can actually split it into 2 pieces. It's using a technology called split-in teams. And what happens is you put each half into a different AAV, you co-transect. And then as they are expressed at the peptide level, the 2 halves then rejoin and form a single editor. And then the editing occurs from there. And this actually works very efficiently. And so we're exploiting this in the eye as well as in the CNS and other places where AAV can work. So initially, our first program is targeting Stargardt disease. So we are looking at the most prevalent point mutation in Stargardt where we're going to directly correct it again, back to normal. We think there are over 5,000 patients with a single mutation in the U.S. And here, you can see dramatic editing efficiency of 75%. This is using that split AAV system to deliver the editor, it's delivering into retinal cells as delivering using a dose of the AAV that is comparable to what Spark has used externa or other AAV retinal applications. So this again would be well in excess of what would be required to arrest this disease and prevent any further loss of sight, which we expect to be around 12% to 20%. So very excited about this program. There is also a recent publication in Nature from David Liu's lab showing use of this AAV delivery system to do a base edit on progeria, targeting the point mutation across progeria. And so again, we see very exciting applications emerging, and we see a lot of applicability of this kind of delivery technology to other tissues outside of [ delivery ]. On Slide 32, just to talk a little bit about our strategy. So having established this platform where we have all of these different components under one roof, it gives us a lot of flexibility. So sometimes, we will, of course, choose to develop medicines ourselves. And take them all the way. That's truly the case of sickle or the T cell malignancies or liver, Alpha-1. But equally, we can do partnerships. And so we pride ourselves to be creative here. So we have done what we call innovator-innovator partnerships something like Verve Therapeutics, where Verve is really an expert company in cardiology and wants to use tools like base editing to permanently lower cholesterol and prevent heart attacks. And so we did a very creative deal with them where we gave them base editing and some of our other technology. And in return, we'll have a share in the U.S. rights in their program, and they're making incredible progress. We want to do more program deals like that where we continue to work with innovators who can use the suite of technology we've built and exploit it even beyond what we can do ourselves. Magenta, a very important relationship we have, where we're working on next-generation conditioning regimens, again to improve transplant options, and expand the use of our curative therapies for more patients within the hematology field. So wrapping up just to say a little bit about the team. It's an amazing group of people united by this mission. We take a lot of pride in the fact that these are people who have worked on novel modalities before, filed many INDs and have delivered approved products. For us, the finish line here is to get new class of medicines approved at the FDA, and have a sustainable engine for new precision medicines that we can move forward, both near-term but also for the long term as this engineering campaign continues. So with that, let me thank you very much for your time. And I couldn't say more about how excited we are about where we're going both in terms of the pipeline and where this platform can be expanded over the long-term to create nuances for patients. So with that, I'll pass it over to Eric, and we can maybe have a couple of questions.

That's great. Thanks, John, for that presentation. Maybe just a couple of questions to start on the sickle program. So the SCD programs, which is really to -- I think the editing efficiency that you guys demonstrated taking a base editing approach compared to CRISPR is sort of well laid out. I'm just curious about sort of the opportunities for clinical differentiation with this approach compared to the CRISPR-mediated assets. What we've seen so far present some pretty high clinical comps, no patient numbers. But I'm curious to know whether at least maybe perhaps looking beyond F levels, there are avenues for differentiation with your asset compared to the earlier -- the first movers using CRISPR?

For sure. So I think at the end of the day, first of all, we're so excited about where the field is, and it's moving really quickly. And it shows us that this is likely to work, and we have a high probability to success, and that's really encouraging. There's no question that editing efficiency being higher, the F level being higher. These are proxies. These are going to be biochemical results in the cell product that we expect it to show differentiation over the long-term in the clinic. There's no question that eradicating VOCs has been very successful for Bluebird and for CRISPR. Nonetheless, I think there's definite opportunity for improvement. We're watching markers of hemolysis, have been really eradicated that underlying hemolytic event with cells that are still sickling, time to engraftment, how viable are these cells and how gentle has your editing process been and how quickly can we get those cells engrafted. Patients are very fragile in that transplant setting. And ultimately, longevity and durability of these edits. I'd say the next-generation as well is going to be to move beyond just these readily measurable markers as well. So there's a big movement in the field to go beyond this crises. Of course, the VOCs will be an approvable endpoint. That's great for the field. And we'll take advantage of that. But ultimately, the goal here is to cure and to change mortality. And we think that there's more to that than just pain crisis. So ultimately, that's going to be about organ damage, and there's going to be an effort in the field to go deeper and to be able to show some of those deeper markers of the disease impact. And so that's work that we're going to do with the field and again, given in this -- what we think is superior nature of the editing that we're doing, we're confident that we will be able to show some of those clinical outcomes as well. The final thing I would say is that remember that there's more going on here than just editing. So ultimately, it's about the payload that you're delivering to these cells and the editing outcomes, of course. But equally, it's going to be about change, will we deliver these programs. And so I think the field is also going to go through some generations where we're all going to begin on this busulfan conditioning regimen, which we know is efficacious and can work, and that's going to provide us a path to approval. But ultimately, we're interested in improving that regimen with non-ablative conditioning and other delivery avenues to make this a less toxic and more accessible regimen. So for us, where we want to position ourselves ultimately in the field is to provide both that best-in-class edit, cleanest, most precise non-cutting, nonviral, highly efficient with the superior regimen for delivery and conditioning and transplant. And I think at the end of the day, that product profile is the one that will treat the most patients.

When it comes to conditioning, I guess, is there any part of that can really be proprietary or better served by your product candidate, perhaps driven by the phenotype of cells or maybe the quantity of cells you need to dose with? I'm just kind of curious sort of how proprietary and modified that conditioning regimen can be.

Yes. It is very possible, and we're working hard on that. I think that, that is a belief we have to be honest. So one of the things about this next-generation conditioning is you're going in, you're a little more surgically precise. At the same time, it's not as much of the sledding. So the question will be, can you deeply enough eradicate the old stem cells to create enough room for your new cells to go in when there will be more marrow around, you haven't wiped everything out. So it's going to be a slightly higher bar. I do believe that may have an advantage for base edited cells where you haven't created double-stranded breaks, right? It's, again, that very mild edit where the cells may have a higher viability heading in to then engraft quicker and compete with what may still be there from the condition.

Okay. Got it. Got it. Got it. Maybe just pivoting to the in vivo applications, AAT and glycogen storage disorder. Well, actually, in one of the earlier slides, your -- you've outlined time lines to bring a liponanoparticle formulation to NHPs and would sort of enabled these programs early this year. And we've seen some initial NHP data with a partnered asset and -- right? So I guess, are you planning to move forward with a similar liponanoparticle? And if not, do you need to sort of individually tailor it with the nanoparticles depending on the asset -- the base editor or the disease application that you're pursuing?

Right. So you wouldn't need to individually tailor LNPs to each editor. So once we have an LNP that delivers this kind of construct, an mRNA of the editor plus a guide to hepatocytes, we won't need to recreate that every time. That's a really important feature of these systems where, again, with very minor gene, there is new guide RNA, and you have an entirely new medicine and a new program. That said, the first time you do it, you do need to get to that right formulation. And LNP is a place of great expertise for us. We have people from Moderna, from Novartis, with -- some of which technology went into Intellia. You can look around the field and see that it is possible. So Intellia has made great progress here. Verve is the company I mentioned earlier. They've shown really nice editing results already in primates using base editing. And so you can see that it's possible today. They showed long-term durable follow-up of those edits in the liver of those primates. So we are doing similar things. We actually have a comparable lipid to what Verve has been -- is doing its own formulation work around that, which is proprietary. So yes, so we -- over the course of this year, expect to be able to share some of our primate editing data using these LNPs. And then from there, you can kind of see the editors are ready to go. They just drop into the LNP, and that becomes a potential development candidate. So we ought to be able to unleash a wave of in vivo programs as that LNP engineering work is done. And then maybe the last thing to say is that I think we're also pretty intrigued by the opportunity to take LNPs beyond the liver. They're synthetic, they're easily scalable in manufacturing, you can redose some. I mean there's a lot of advantages there. And we do think that liver is the low-hanging fruit, but there are going to be abilities to move that into other tissues over time, and that's an area we expect to invest in as well.

Okay. Okay. Got it. I guess looking beyond sort of LNP selection, how should we be thinking about the in vivo directed program sort of advancing to final DC selection and IND-enabling studies?

Yes. So I think if we stay on track with LNPs this year and I can share that data, then we've guided that we have at least one development candidate this year. And again, there is really -- we think there is at least 3 in view in various stages of late-stage lead-off. So the primary work here is LNP, that's the purple path. Once we have that, we'd be at that DC stage. And then we expect a pretty predictable path to IND after getting to DC.

Okay. Okay. Great. I think we'll have to leave it there for time. Thanks again, John, for your time this afternoon and getting us up to speed on the story. And thanks, everybody, for tuning into the webcast. Everybody, have a great afternoon.

Great. Thanks, Eric. Thanks, everybody.

Yes. Yes.