Great. Welcome to the next session of our first day of the 2025 Global Healthcare Conference. I'm Mani Foroohar, Senior Analyst here at the drug and medicines team. I have the good fortune of hosting John Evans, CEO of Beam Therapeutics. Thank you very much. John, busy day for you today. Busy day, yes. You guys released early but meaningful AAT data this morning, closed a pretty substantive round, and sort of extended your runway into 2028-ish, if my memory serves. Let's talk a little bit about the relevance of the threshold in AAT that is talked about by every company in the space, which is this 11-micromolar threshold. What that actually means clinically versus where heterozygous humans and sort of healthy, normal humans actually live in terms of serum. Let's understand what that number actually means. Yeah. First of all, thank you. It's great to be here. When you think about AATD, you have an autosomal recessive condition, right? Patients have this one mutation. It's called the Z mutation. It's a single-point mutation in their gene. That causes two problems, right? One is it's a mutant protein. It builds up in the liver, and it gets stuck there. It causes a lot of liver damage. Because it's building up in the liver, it's not secreting to the lung, to the circulation, where it needs to protect your lungs. When you have two copies of that, you're very severely deficient in alpha-1 protein. Your circulating levels might be in the mid-single digits, right? Four to six-eight micromolar of AAT. Your liver is chock-full of the stuff. It's obviously causing liver damage. You end up with this horrible, progressive, inexorable emphysema, literally loss of lung tissue and function, particularly exacerbated in those moments of infection when your body is mounting a defense. The alpha-1 protein isn't there to stop those defenses from attacking the lung tissue. You lose a lot of tissue in that moment. In any autosomal condition, we ask ourselves, what is the threshold for not having the disease anymore, right? Almost always, it is the carrier state, right? The parents of a child who gets one of these diseases are normal. They don't have that disease. When we look to the carriers in this condition, people who have one copy of Z, there's two. There's SZs and MZs. S is also somewhat dysfunctional protein. It's sort of an intermediate, I would say. MZ is a full carrier. You've got one normal copy, one Z copy. These patients do not have disease, right? They don't have alpha-1. They're not diagnosed with it. They are clinically normal. The only thing you can say about them is that if they have a second hit that's really... I think for our initial data set, we're already at the therapeutic threshold. We're at 12.5 and climbing. We can easily push it higher. I think we do intend to. It's not completely clear to me how much clinical benefit we would ever be able to show from here forward. I think we are basically in the therapeutic area. We have transformed these patients' profiles and potentially have a functional cure for them going forward. Let's talk a little bit about disease management and what AAT looks like on the scale of an individual patient. Because AAT is not like some rare diseases, an illness that in most patients results in a rapid decline. It is often a disease that makes you more—it's a disease that, in terms of lung disease, makes you more vulnerable to other insults. Over time, the sort of slings and arrows of breathing air will eventually accumulate damage. Worse for people who live in highly polluted environments, worse for smokers, those with frequent severe infections, et cetera. Let's talk about the clinical impact of turning someone from the phenotype of a ZZ patient to the phenotype of a ZM carrier, which is what you've genotypically achieved. What do those two pictures look like? What level of cost, hospitalization, expense, mortality, what are you actually preventing between those two? Yeah. It's a great question. It kind of illustrates the whole point. If you have the alpha-1 antitrypsin deficiency, not only are you managing your disease, you're declining, right? Now, the rate of decline may be variable in patients. It obviously does depend on when those hits happen. There may be variability there. On the whole, you are on your way to an increasingly severe emphysema where you lose your lung function, an increasingly severe liver failure, often towards the end of life characterized by double lung transplants, living on oxygen from the lung side, and liver failure and liver transplants on the liver side. It ends up being effectively terminal on those organs. It can be fatal, of course, for the patient as well in some circumstances. Certainly, hospitalization-wise, there's a lot there. To manage that, we have some options today. We have chronic augmentation therapy that is used, which is imperfect, I think. It is at least available. Obviously, the cost and hassle of that is something to consider. Compare that then with if we can take someone from that disease phenotype into that MZ carrier status, which is where we think we are. At that point, they should have no more worry of progression, right? Their lungs are going to stay as they are. They are not going to lose any more tissue. That also implies that they are not going to decline towards further and much more expensive care at the end of life, right? Because they are not going to get much worse. Now, will the lung function recover? I think there we have to be more cautious. I think once you lose lung tissue, it's hard to imagine getting it back. At the very least, you would have stopped the progression. You've frozen the disease in place. Liver is a little different. Liver, I think we actually have some reasonable expectation that not only would you relieve and block the progression in the liver, but you'd actually start to relieve the organ itself. We know that from a variety of different preclinical and clinical data sources. Clearly, if you can stop the production of this constantly aggravating toxic protein in the liver, basically the liver can heal itself. The liver is very regenerative. I think the liver, we would expect recovery of function over time. I think lungs, you're trying to just stop the inexorable decline and preserve what function the patients have left. All of that is with a single therapy, right? One regimen. You permanently change your AAT levels. You transform yourself for the long term. Obviously, your adherence will be perfect. You'll never miss a dose because your liver is doing the job now. We've literally permanently given people a reprieve from this disease. When we talk about the path to drug, from data set to drug, it gets a little more complicated. I know we know a little bit about what could be one pivotal design from an oligo approach for liver disease by another company, Arrowhead, partnered with Takeda. Ongoing. We'll see what that data looks like. How do you think about the path to commercialization, the provability in liver versus lung for a one-time therapy as opposed to sort of a chronically dosed injectable? I would actually spin it a different way. I actually think we have an embarrassment of riches when you think about how to move this program forward. The beauty of fixing the disease at the root cause is you know that all the downstream consequences of the mutation are going to play in your favor. That is exactly what today's data set showed, right? We edit the one root cause of the disease, the DNA. Sure enough, you now start producing M. M goes up dramatically. That, of course, drives your total AAT up above the therapeutic threshold. At the same time, you are reducing Z. We are literally going to stop producing as much of the toxic protein that is causing the problem. All of the AAT we produce is functional, OK? At the highest dose level, nearly all of 88% in the last data point of the circulating protein is M, and it's functional, OK? You're dramatically changing it. What that means is we have all of those endpoints available to us for development of the drug. When you address the full spectrum of the entire disease, you can go in a lot of different directions. I would say there's basically three different kinds of approaches we can take to developing the drug from here. The first and foremost is actually just to consider an accelerated path to market. I think that these precision medicines with very strong science, when you're right on the mechanism of the disease and you have biomarkers that are all going in the right direction in a coherent way and are predictive of clinical benefit, that is exactly the setup for something like an accelerated approval, OK? That is the conversation we're going to have to have with regulators, of course. You want to bring them strong data in order to have that conversation. We do think that this is a plausible path. In that case, it could be as simple as expanding this current trial until we have enough N. We've obviously finalized whatever the dose may end up being and then go forward from there. I think that's a very attractive and obviously potentially fast path. From there, then we also can consider studying lung endpoints, right, in terms of that point about arresting progression of the disease. We think about things like CT densitometry, where you're literally assessing the density of the tissue and comparing it over time and trying to show that you've slowed that or arrested that progression. Liver, as you know. We would have beneficial effects across all of those more functional organ endpoints. We can use any of them and probably will use all of them to show the effect of the drug over time. In the interim, we may do a little bit of work, for instance, in expansion phases of the Phase I, II, where in the lung patients, we may do bronchoscopies to sort of get into the tissue and show what's the inflammation level, what's happening there locally. In the liver cohort, you can do liver biopsies as well. OK. I'm going to move over to the practical patient experience, if that's all right. The patient experience in the real world currently of protein supplementation is a little bit on the onerous side. Frequent injections, going into the physician's office, travel can be a challenge for some patients. That varies depending on how far they live from an urban center or infusion center. Walk us through the process of administration and follow-up through the single dose of BEAM-302. Yep. I think you gave away the answer with the word single. I mean, I think that's the whole point, is this is a single administration for a lifetime of benefit. That has huge implications, obviously, for patient convenience and quality of life, for caregivers and family in terms of their support for this, not to mention, by the way, payers and the health care system, right? A lot less utilization of health care resources in terms of delivering care for these patients from a go forward basis. We think that is an ideal sort of option. When you talk to patients with this disease, you hear that loud and clear. They are sick of the kind of chronic regimen, the augmentation therapy, if they even have access to it. If they're in other countries around the world, they literally don't even have that. They're just desperate for an option. There's really very little. It is very inconvenient. Contrast that with here, with BEAM-302. We're doing an LNP dose. It's lipid nanoparticle. It is as simple as a little bit of steroid pretreatment, which is standard in the field from Onpattro forward, and then a few-hour infusion in the chair. It would most likely be an outpatient procedure. You'd go home afterwards. There's no monitoring. There's no intervention needed after that. From an efficacy perspective, what we've already shown in this data set is just amazing. You literally, within a few days, start to create the new protein. The changes start to take effect. We know from preclinical studies that the base editing equipment is probably gone from the body within several days, certainly by a week. It is all out of there. The genotype has then been corrected. It is just a matter of the system basically recalibrating over a few weeks. You reach a kind of new plateau. It is very fast. Certainly, within a month, you have basically changed completely the profile of the patient. You have taken them from disease to non-disease. I'm going to put on the hat of some of my former colleagues in hedge fund land. John, if you're giving a lifetime benefit, but in the U.S., commercial patients are only on insurance for a defined number of time, and they change between plans, and insurance plans manage their budgets yearly, how are you going to get paid for lifetime benefit? This is a question not just for you, but one that gets asked across the genetic medicines universe for one-time agents across the board. How do you answer that? The opportunity is not what you think it is, sort of bare thesis, specifically in the case of AAT and for BEAM-302. Yeah, it's a great question. Look, we face that a lot. I think a few things to say. I mean, it is a novel pricing regime that we're moving into. We have to answer some of these questions. There are some puts and takes here. I mean, on the one hand, the pharmacoeconomic argument for the prices that we're going to be talking about for these sorts of gene therapies is easy to make, right, even at those higher premium prices because of three things. We're treating patients in the prime of their lives. We're going to give them a lifetime of benefit, quality years, productive years. That counts a lot in the pharmacoeconomic analysis. The second is we are going to displace a lot of medical care in the future, hospitalizations, transplants, all of these sorts of things. We're going to displace, to a certain degree, chronic medicines, right, other pharmaceutical and biotech products, which are expensive, as we know. They last for life. A one-time cost, even if it's on the high side, relatively quickly is paid back because you're saving the whole system money and you're providing a lot of value to the patients. I think that's sort of an important feature, I think, of what we're doing. The question of the lumpiness of the payment, right? It's a one-time check. Somebody has to write it. Yes, patients will shift from plan- to- plan. We do need to overcome that. Of course, if you have broad coverage across the country, which so far we've seen happen for the various gene therapies that are being approved, it will balance out, right? Plans may treat someone and lose them. They're going to get somebody who was treated. As long as the patients are all still within the net system, it should actually average out. There are more creative types of things we could explore, whether it be amortization of some of the payments or sort of agreements with the plans where there's some predictability on the price. I mean, some of these things have been explored. Things like Zolgensma, right? It's been an effective product, effective launch, high price. I think when the drug really works and there's a strong demand for it, I think that these sorts of issues can be solved with the payers. I think one piece of pushback that I get is, but to some extent, can't AAT patients marginally reduce the impact of the disease through their lifestyle? If you don't smoke, if you don't live in a high place, highly polluted, et cetera, can't that resolve it? Right. The answer is no, I mean, very clearly. You definitely should not smoke. I mean, if you are an AAT patient and you are smoking, you are crazy. You will fall off a cliff. In the absence of smoke, never smokers with AATD are losing function constantly, right? That is why they are diagnosed, because they are going to show up sick. They look like smokers. In fact, one of the stories in this disease is diagnosis. We have got to go find some of these patients. There is a large number of diagnosed patients, but some are undiagnosed. The undiagnosed patients, it is not like they are OK. They are just living in COPD clinics. They tell their doctors they have lost breath, they have lost function, they are struggling. The doctors say, are you a smoker? Sort of that suspicion comes in. The epi is very clear that you do not have to smoke. You can live a perfect life. This disease will take you down anyway. Let's talk about the path forward from here in terms of data. Obviously, we're going to be seeing enrollment of patients with lung and liver disease going forward. Talk to us a little bit about the patients coming off of protein supplementation therapy and how you manage that washout period, just to make sure you get a clean data set. What the next couple of data sets in each patient population will be between now through to next year. Yeah, totally. Just as you said, the next steps in the trial, we're going to continue on with the dose escalation in the lung, the part A. That will be, I shouldn't say in the lung. It's just mostly in patients who don't have really heavy liver involvement, just AATD broadly. We'll have another dose up here. We know we're in the steep part of the dose escalation curve. We're excited to see where that takes us. Of course, we've been very well tolerated. We think we've got a lot of flexibility in where to take the drug from here. That certainly is a data set that will mature in the near term as we continue to add patients and find that optimal biological dose. We have guided to at least another data update in the second half of the year, which I think would most certainly cover that information. We are also going to be kicking off this liver cohort. Again, as a reminder, that is a minority of the patient population. That is where people who have really stressed livers as almost a predominant feature of the disease. That is a minority of the population. Maybe it is 15%-20% of total. It is important to check that. There, we are going to do a relatively efficient, I think, redose escalation just to confirm. It may not even involve going all the way back through all of the doses we have already done. That will be important data just to confirm there is no major difference in safety and efficacy in those patients. Our base case understanding is there would not be, especially based on the modest nature of any liver signal we have seen in our LNP dosing to date, that we think gives us that window quite clearly. The goal is actually to kind of bring the populations back together so that we are not talking about lung and liver. We are talking about a single population across the whole spectrum of disease. That is the vision we have for the drug. I do not know how fast we will have that data. That could be part of the second half update. It could be a little bit after that, certainly into next year. I mean, at this point in developing a drug like this, we are going to do it as fast as we can. We are going to add a lot of N now to the trial. Some of them, patients will be answering some of these specific questions. At some macro level, you're just trying to enroll a lot of patients. You build up your understanding of the drug, your database. That's the substrate to have some of these conversations with regulators about where we go next. I actually think the trial is going to accelerate from here over the next year. Those data updates will be fairly meaningful. One of the things that we've talked about and that one of your competitors often talks about is M versus not quite M, or depending on who you are, use a different word for it. We'll call it bystander, which is M corrected but with an additional edit, which I think is the most neutral description one can provide. You gave numbers this morning in a more detailed presentation on the conference call at 8:00 A.M., digging into the amount of total AAT you're producing, but also functional metrics of neutrophil elastase. Talk to us a little bit about the contribution to neutrophil elastase inhibition or neutrophil elastase binding, depending on which assay you want to use. It It doesn't actually matter, of M protein in this sort of vanilla sense of a homozygous non-mutant patient versus M protein with an additional mutant. We will not talk M protein with an additional edit, discuss neutrally. Yeah, yeah, it's great. I think you're exactly right. When you talk about the disease, right, this is not a disease that's caused by mutations all over the gene, which some diseases are. This is a disease where there's a very specific point mutation, which is the Z allele. It's at a very specific position in the gene. We need to fix that. If you fix it, you get the M allele, right? That's the position that defines Z versus M. BEAM-302 does that, right? As you note, one of the features of base editing, it's rare, but it can happen, is in this case, we can occasionally make an additional edit at a neighboring allele, OK? You will sometimes do that, in which case you will get what we call an M variant as opposed to the canonical M, right? Canonical is sort of like in the literature, what's the sequence? The reason we call it a variant is actually there's a lot of evidence that humans have variation at that site, OK? There's actually a large number of variants at that site. The one that we create has been seen in people, so it has human validation. There are others. It shows you that that site is tolerant of change, OK? Once you know that, you just do the work to characterize the protein. It's almost like the fingerprint of our edit. We had to convince ourselves years ago that that was fully active and behaving exactly like the canonical M. I think that preclinical data at this point is overwhelmingly clear that that is the case. We can go into that. I think we've published a lot of that previously. Nonetheless, really exciting then to show the data today that shows that all of the M that we're creating, which is likely a mix of M and M variant, is functional. Again, that's exactly what the preclinical data would have suggested. If you think about it, that's what really matters, right? One is sort of the biochemical levels you're achieving. You can think about M and variant. At the end of the day, it's how much functionality do you have there? Patients with, and this is another important point, patients with the disease, they have 100% Z at baseline, right? Z is not a good protein. It's bad because it builds up your liver. It is causing liver toxicity. Even in circulation, it isn't a good protein because, one, it's less effective. It's lower on the functional assays. It also polymerizes. We think those Z polymers can cause inflammation. It disrupts the activity that is supposed to be happening, the protection, OK? Even in the context of augmentation, you can put augmentation in. The Z can get in the way, OK? Augmentation does not get rid of Z. You want to get rid of Z. What we're doing is not only are we raising total and we're above the therapeutic threshold, by the third cohort, it's almost all M, right? We're 88% detectable M in the periphery versus Z. That means of the 12.5, if you think about it, we would be about 1 micromolar of Z. The rest is all M, right? That is a major change in what is actually circulating in the body. Finally, we showed that you measured the neutrophil elastase inhibition, which is the gold standard assay here, of how much is this AT actually inhibiting neutrophil elastase, doing its job in the body. We showed that it is. In fact, it is at the same levels as you have total AT. Those lines converge because we have driven away all the less functional Z. All of our protein that we are creating is fully functional. That is inclusive of M and M variant. I think that data set is pretty definitive at this point, that we have a fully functional editing outcome and that we're restoring not just total levels, but also functional levels of AAT in the body that should be fully therapeutic. Let's talk a little bit about the opportunity set and the TAM. And how you think about where you should be accessing it versus where, as you move into larger studies, larger than 10 patients, where it's appropriate to seek a partner. This is obviously a large, fairly sophisticated global market. There exists a fairly rich commercial market for protein supplementation, not in all countries. Some places it's not paid for. But talk to me a little bit how you think about global strategy. What is core to Beam? What is sort of not core to Beam? Yeah, great question. We have two focus areas in general: hematology, which of course has sickle cell disease, and then liver with alpha-1. I think of the two, we have the capability of moving them both forward under our own steam. Hematology is clearly a place where there's a lot of complexity there. There is an open conversation about do we keep it? Do we get some help? We'll see over time. I think in liver, there are easy supply chain. It's LNP. It's very straightforward to think about commercializing. There, I think you have a higher innate likelihood of saying, just strategically, we can do this, OK? I would also note that with alpha-1, we talked about the pricing in the U.S. should be gene therapy pricing, right, which over tens of thousands of patients is a very large TAM and with not a lot of other options for those patients. Ex-U.S. actually is pretty good too. Ex-U.S. countries in Europe, they actually really see the need for this population. Of course, there, they do not even have augmentation therapy. I actually think, unlike some gene therapies where there is a big drop-off, I think there is significant global commercial potential here. Will we eventually need help? We will see. This could easily be a TTR for us, like an Alnylam, which they use to go global. I would not close the door to doing that. Certainly, the existing patient population, our ability to treat them with a drug that would be transformative, I think that would be a successful launch scenario. Some of the other things, like going global, driving diagnosis if you're reaching deep into the primary care networks within the U.S., those are places where we'll have to decide, are we going to build that ourselves? At some point, do we get help? I think that's pretty far down the road. We're running a little bit late. One more comment on the European question on the European market opportunity. Obviously, as you mentioned, there's augmentation therapy not really reimbursed there. Do you think you will have to show event rate benefit, FEV, some kind of more hard endpoint to be reimbursed in Europe as augmentation therapy was? What might that look like? Yeah, I think that goes into the whole clinical development plan. I do think that as much as I think this is a perfect setup for accelerated approval, we will need to show those sorts of functional benefit endpoints over time, right? That's just part of doing good drug development. It's good not only for regulators, but it's also good for payers and driving that value story and, frankly, creating an insurmountable lead with competitors who will eventually try to come into this market. This is a disease indication that is easily big enough to justify those kinds of investments and goes back to the basic confidence we have in the drug. If we're really correcting the disease at its root cause, we have high confidence that anywhere else we look, whatever endpoints we want to measure, we should see the move in the right direction because we have fixed the disease at its source. Great. On that, we'll wrap up. John, thanks. Congratulations again. Thank you very much.