Welcome to our first in-person Global Healthcare Conference post-pandemic. It is really nice to see everyone. I would like to thank all the participants, investors, companies, also especially our event team and corporate access team who made this event possible. With that, I would like to introduce our next presenting company, Beam Therapeutics. With us today we have Pino Ciaramella, Chief Scientific Officer and President. Pino, maybe before I introduce, could you give a brief overview? Of course. About Beam? First of all, thanks for the invitation. Finally, to see people alive rather than through Zoom is a wonderful experience again. My name is, as you can see, it's Giuseppe Ciaramella, but everybody calls me Pino. It's an Italian thing, but please call me Pino. I'm the President and CSO at Beam. Beam Therapeutics is a next generation genetic technology that primarily is based on a technology that was originally developed by Professor David Liu at Harvard and now at the Broad Institute, which is called base editing. Base editing is still a CRISPR-based gene editing technology, but with two fundamental changes. The first one is that the CRISPR protein is modified such that it can no longer make a double-strand break, which we believe is a significant stress to the cell. In fact, the protein has then been further modified by attaching to it a human enzyme called the deaminase, which is capable of catalyzing a direct chemical reaction, which is called a deamination, directly in the gene, and can convert one nucleotide to another simply by eliminating essentially an amine group from the binding surface of the nucleotide. Depending on the deaminase that we use, we can actually catalyze a C to T change or an A to G change. As you know, a C or an A are present in all of the base pairs in the human genome, and therefore, if you target the leading or the lagging strand, you actually can achieve four edits with the two editors. Of course, there are many things we can do with this editor. Of course, we can correct direct point mutation that cause disease back to a wild type sequence. Actually, because base pairs carry essentially many of the different biological functions, we can change other things as well. We can change regulatory elements like we are doing in our BEAM-101 program, whereby in this case we are actually inserting point mutations in the promoter region that prevent the binding of repressor proteins. We can change amino acids in the active sites of enzymes, making them more or less powerful. We can change phosphorylation sites, therefore modulating regulated sort of signaling cascade. Then finally, we can also introduce stop codons, as well as disrupting splice acceptor donor sites so that we can knock out genes very effectively. It's a very versatile tool, and we've now deployed it across more than 12 programs in the first four years or so that the company has been active. Thank you. Your lead candidate, BEAM-101, will enter clinic pretty soon. What is your thoughts on, you know, the initial phase I trial design? Yeah. What is that? In this case, in BEAM-101, as I mentioned, we are modifying the promoter of the HBG1 and HBG2 genes. These are the two genes that are responsible for expressing the fetal form of hemoglobin, which in adulthood, soon after birth, are typically switched off by the binding of repressor proteins. By making those point mutations, those repressor proteins can no longer bind to it, and at the end of last year, we filed a successful IND, which is now open, and we plan to basically select the first patients in the second half of this year. The trial design will be very similar to probably what you've seen with CRISPR Therapeutics. It is U.S.-based, at least initially, and it will require what is called a sentinel cohort, which is a cohort of three patients who are sequentially dosed. The first patient is selected and dosed, and then we need to wait for engraftment to take hold, making sure that there are no safety concern. Then at that point we'll dose the second patient and the same process will be repeated, including the third one. At the end of the third patient, we will then have an expansion cohort, and a subsequent two expansion cohorts that have the potential to become fileable, you know, pivotal, particularly in the last cohort. We will see obviously how CRISPR is able to actually file. They've mentioned that they'll probably be filing at the end of this year. We will see whether we can follow that same similar pattern. Okay. Can you give us a sense at what point we will be able to see the initial data and after how many patients? Yeah. We haven't frankly neither guided nor decided to some extent whether we want to disclose just a single patient or whether we want to have the opportunity to have evaluated a number of patients, in particular, have followed them for a certain period of time. I think remains to be seen. The important thing to think about as we are upregulating the fetal form of hemoglobin just the procedure in itself can activate and elevate the fetal form. It's called erythropoietic stress. It's important to wait at least six months in order to ensure that that stress is calmed down and the results that you see are truly attributable to the therapy rather than the procedure itself. That's why we are not yet committing to a rapid disclosure of data until we have had the opportunity to evaluate the circumstances. We will let you know as the program progresses. Is it fair to say, like sometime 2023 we will be able to see the data? Well, it's fair to say that it could be possible sometime later in 2023, but whether we will be willing to do that or not, as I said, depends on the totality of the data. We really want to provide a single patient data, if we can. Second half this year. Mm-hmm. Several questions, why it will take so long? The second question is, you know, given the experience of non-transfusion dependent anemia in two patients and how would you thinking about patient selection? Yeah, I'll come back to the second part. The first part, the timing is frankly a technical timing. These are, they're essentially transplants. You know, that's the therapy. It requires, first of all, to select centers that have experience of that, and we have already selected several in U.S. You need to go through ethical committee as well as IRB, and contract that typically takes about six months also just in that process. You then are allowed to start screening the patients. The patients that have been selected will need to undergo at least one month period of flushing out hydroxyurea from their system. They're typically chronically treated with a hydroxyurea, and that's not something that we want to have on the study. They will have to receive transfusion during that period of time. They will then need to come back to mobilize their cells from the bone marrow in order for that collection of the cells to take place. Typically, they receive at least two mobilization in order to generate sufficient cell. Then at that point, those cells will be manufactured and eventually the patient will be conditioned in order to create essentially the niche for the edited cells within grafting. You can see how this is a pretty long process in order to do that. The good news, though, is that once you start the clinical trial, actually the potential path to pivotal and licensure is actually very fast. If you look at CRISPR, it will be roughly about five years from starting of the clinical path, and we plan to follow similar timelines as well. In terms of your questions around the bluebird bio experience, we are not planning currently to genotype for fetal hemoglobin. However, we do collect samples at baseline prior to treatment for all the patients, and we will have the opportunity to evaluate whether something obviously is, you know, occurs like the anemia that has been seen in the beta- globin bluebird trial. The other thing to say is that we would exclude any patient that would have had thalassemia prior to the study. Okay. Another very technical question. When you say you know, I mean, you did mention you were editing both promoters of HBG1, HBG2. Mm-hmm. You know, be it in the pre-clinical or the human cell lines, you show like editing efficiency, like over 80%. Can you know, give a little more color? Were those the monoallelic, biallelic, and you also have four sites, and when you say over 80%, under what context? Actually, for BEAM-101, typically, we see more than 90%. It's with BEAM-101 or BEAM-102 that we see about 80%-85%. We have done single cell studies, although these have some technical challenges because you need to differentiate the cells. From those studies, it's clear that more than 90% of the cells actually are quadruple edited. All four loci are changed to the permutation that we want. We have about 65% of cells which have two to three edits. They're really very robust. You would expect that once the editor actually goes into the cell, it's capable actually of making the edits on all of the four loci as part of that. We measure the hemoglobin S activation. In vitro you need to differentiate, and even that procedure itself can cause some changes in the fetal levels itself. It's not the most reliable of study, but we consistently see very high level of hemoglobin S activation. In engraftment studies in the mouse, where I would argue it's probably the more relevant and reliable data points in terms of a upregulation of F, we see the highest level of upregulation that anybody has reported, typically 60%-65% upregulation in cells isolated from the bone marrow of mice after 16 weeks of engraftment, which is the only time at which these stem cells have actually now are being repopulated bone marrow. The concomitant decline of hemoglobin S as you reach those very high level of S is probably equally clinically relevant. We typically see less than 4%, which is the levels that I've seen in sickle trait individuals. These are individuals who are heterozygous and do not have symptoms of sickle cell disease. We believe that with that approach, we're actually getting almost like two for one highest level of F, which out-competes the S protein as well as the concomitant decline of S. Nobody has reported that, not even the clinical data that have recently been disclosed through the CRISPR study, there is still about 50%-55% hemoglobin F, which, as you know, is above the sickle trait individuals. Okay. Would that, you know, highlight the differentiation regarding the promoter area approach of HBG2 versus BCL11A enhancer approach? Yeah. I would say obviously the highest level of hemoglobin F as well as the decline to hemoglobin S, we do believe that the higher that level, the better it is the clinical outcome. Of course, you can start to see some benefit below that level, but actually there are many parameters that can hopefully more deeply and more quickly resolve that ultimately contribute to what is sickle cell disease. As you know, there is a very noticeable and obviously impactful as far as the quality of life of individuals, is this pain crisis that the sickle cell patients actually suffer from. They're called the vaso-occlusive crisis. It looks from the CRISPR data and the bluebird data that those are relatively easy to fix. In fact, they get about more than 90% improvements over the prior 12 months or so. However, really, these patients are prematurely killed not by vaso-occlusive crisis, but by the progressive damage to many organs, including the brain, typically the lung and the kidney. It's this progressive damage that we are looking to at least stop, if not regress, in order to achieve a curative therapy for sickle cell disease. As part of our clinical trials, we will monitor many of those parameters that contribute to that progressive organ damage. Typically, this is in an inflammatory milieu, which is generated by the fact that the blood cells in sickle cell individuals lyse very frequently. We will look at the half-life recovery. We will look at also at the hyperproliferative consequence of the fact that these cells are essentially lyse very frequently. This will contribute also to aspects of the quality of the blood, like the viscosity, for instance, the prevention and removal of hemolysis. All of these factors will be able to give us a clinical picture which is much more complex than just VOC, and will hopefully tell us that we are on the way to prevent this progressive organ damage. For you to share the initial data, what would be the threshold for, you know, the numbers you quoted? Yeah, as I said, we haven't decided. It's probably not even more the numbers, but it's more the follow-up time. Mm-hmm. As I mentioned before, we want to make sure that what we have seen, particularly the upregulation of hemoglobin F, is not the consequence of the transplant itself. It's not a consequence of the stress that these cells would undergo as part of the process, but actually it's truly attributable to the therapy. It's really more about how many patients we have with a reasonable progress with a post-transplant that gives us confidence. I'm sorry. Gives us confidence. That's the Italian in me moving my hand. Gives me confidence that basically we actually have a therapy that's doing what it's supposed to be doing. You also have another program, the Makassar program. Mm-hmm. In the sickle cell. Maybe can you lay out how do you see the initial program differentiate from, let's say, CRISPR's approach and how the second program will have a further improvement? Yeah. In the Makassar approach, which we call BEAM-102, we are taking the ability of the base editor and its precision actually to convert the single point mutation that's causing sickling into a natural, naturally occurring normal variant called the hemoglobin G-Makassar. This approach actually is moving slightly behind BEAM-101, but making very good progress. We love the fact that we are uniquely placed to actually do a clinical study, the upregulation of hemoglobin F, where we do not completely eliminate the sickling form of hemoglobin, but you are reducing it significantly and you are outcompeting it with hemoglobin F versus the actual elimination and the curing of the cells from hemoglobin F itself, where in that case we're directly converting F into a normal form. We... The plan is to essentially take them both to clinical studies and then, frankly compare the clinical data between the two mechanisms and see which one has the better impact on the organ damage progression. That, as I said, is our aspiration to generate a cure, treatment for sickle cell disease. We'll probably likely take two pivotal trials, only one of the two approaches, in order to move forward. The other point to make though is that BEAM-101 with the hemoglobin F upregulation also has the ability to treat beta thalassemia. There may be a differentiation also at the level of the indication that we actually can pursue. I think quite some pushback from investor regarding sickle cell disease indication is it's such a crowded space. You have several candidates ahead of you, and they are setting pretty high bar. How much do you really think, you know, the differentiation, the additional benefit you can provide in order to have some meaningful market opportunity there? Yeah. As I said, you know, the good news for sickle cell disease patients somewhat is that even though unfortunately they do die prematurely, the disease is not a near acute killing and therefore it allows the opportunity for these patients basically to be warehoused and wait for the best possible treatment that they can actually receive. As I mentioned earlier, we do believe that the highest level of re-regulation of that or actually the complete elimination of pretty much of the sickling form of hemoglobin are going to provide the best clinical outcome for these patients. The treatment here is about curing, and it's not just about fixing VOCs without having a positive impact on stopping the progression of the organ damage. That's why we believe that actually even under the initial paradigm, which will utilize busulfan as the conditioning agent, we have two potential very best-in-class sickle cell disease. We do believe that through that package of clinical information that I mentioned earlier, that we will be able to demonstrate the ability of this deeper resolution of sickle cell to have a positive clinical impact on patients. Having said that, though, we do think that the patient population for sickle cell disease has the ability to expand very significantly as there is further progress in this treatment paradigm that moves forward. We've recently articulated that we see this treatment paradigm evolving in three ways. Wave one is the one that we've just mentioned here, which is essentially a transplant, which is aided by busulfan, which unfortunately, even though it is the standard of care, it is a somewhat toxic regimen and leads to sterility. We and others are actually actively developing improving conditioning regimens that do not cause sterility and can frankly increase the patient population that can actually and is willing to undertake this treatment paradigm. Eventually, we are developing and we're innovating significantly in delivery technology, including lipid nanoparticles that actually are capable of going directly into the bone marrow and the stem cells in the bone marrow, therefore completely potentially eliminating the need for transplant. That would be wave three. Once we achieve that paradigm, we do think that we have not only now it's against a population that is much, much wider than the current population exists. Typically right now is about 10% of the sickle cell patients are the ones which are eligible for the transplant. By going through the two and three, we see a very significant expansion of that patient population. We can be competitive both in terms of the best in class as well as having potentially a very significant portion of an increasingly large market. Thank you. We have a few more minutes left. I wanted to touch on in vivo approach. You do have a partner Verve already have a lead indication in targeting PCSK9. You also mentioned identify your own first one that's a GSD1a going after R83C mutation. Maybe give us a little bit more color that what is the main reason choosing GSD1a to a specific target? What is the limiting factor there? Is that the editing efficiency in vivo that level of the protein that you require to be generated? You know, the choice of GSD1a is simply driven by the fact that it is a very significant medical need, and base editing can essentially correct the point mutations causing the disease back to wild type. We have shown in transgenic mouse models a very significant correction of not only at a high level of editing efficiency in heterozygous animals, we are typically seeing 60% correction, which actually in the liver is getting close to saturating levels of editing because hepatocyte, which is where this editor is targeted, is about 60%-70% of the entire liver volume as part of that. Obviously that's not the only one that we have. We have two other programs. One is actually the correction of the single point gene that causes alpha-1 antitrypsin, E342K mutation, and that's also progressing well in preclinical study, as well as a second mutation for GSD1a, which causes the premature stop codon at Q347X. At Beam. The way we prioritize our portfolio is we are actually prioritizing strategic buckets, and then each one of those strategic buckets as the resource to move at speed at the pace that they can. Studies they will just continue at the pace. There is a difference between, let's say, alpha-1 and GSD1a. GSD1a has a lower bar in terms of editing efficiency, but needed, you know, different transgenic animal models to be able to understand the disease. The biomarkers are more complex. Alpha-1 actually has a biomarker, which is essentially the protein that one can look at very readily, but it needs a little bit more of an editing efficiency. In both cases, actually, we are achieving that in preclinical animal model. What is the more challenging aspect of correction as opposed to knockouts, which is what you see in PCSK9, is really how you predict the clinical starting dose. Because the translation between a mouse model versus human is not necessarily straightforward, as for a knockout. So really what the preclinical package focuses on is creating a PKPD relationship between mouse, non-human primates and hopefully human that allows us to a good prediction on the starting human dose. That's basically what we're in the process of doing. Okay. In 2022, you will identify or select a second candidate. Mm-hmm. Liver-targeted. Any color you can share, is AATD a good guess or it would not be A- It will be either. Okay. It's alpha-1 and GSD1a, both of them are moving frankly at the same pace. We're not prepared to say which one is gonna happen quick. We'll just, you know, push them at the speed that they can go. Of course, I would also remind you that we have two partnerships, one with Pfizer, the other one with Apellis, also with some liver targets, and those are making good progress as well. Since you mentioned Pfizer partnership, how's the progress with muscle and CNS targeted or directed lipid nanoparticle development? We've chosen three targets. We know exactly what the targets are, although they have not been disclosed or valued. This is up to Pfizer to disclose what the targets are. One is for the liver, one is for CNS, and the other one is for muscle. For the liver, obviously, we have the LNP technology is already there. For the muscle and the CNS part of the work is actually to discover these LNPs that can go to those tissues. Thanks to the Guide technology that we acquired last year, we can actually barcode each individual formulation. Literally, those hundreds of this formulation in the same animal at the same time. At the end of the study, we can look at the different organs, and the barcode will tell us which formulation is actually going to which tissue. We have a rapid way of developing LNP technology that can actually have the biodistribution that we really are after. That's in part why Pfizer was interested in creating basically this relationship such that it would stimulate further discovery work in CNS and muscle LNP, and that's what we actively do. Thank you. Thank you, Pino. We are running out of time. Thank you. Thank you very much. Okay. Thank you, everyone.