#409 ‒ Inside modern drug development: the science, economics, and regulatory hurdles behind bringing new medicines to patients | Lloyd Klickstein, M.D., Ph.D.

24:29Peter Attia:Which, to your point, by the way, nobody listening to this doesn't care about this. So this is a very topical consideration. It's not esoteric. It's something I've personally been working on for 20 years, as you know. So what should be the drug format? Well, the patient population are likely to be older adults. So it needs to be a drug format that'll be suitable to them. And shouldn't underestimate the importance of this. It's what's most likely to work. because in a drug development program, especially with a new target and a new indication, there are so many unknowns and all of the risks multiply across a drug development program.

25:10So if you have, say, 20 risks you're taking in a drug development program, whether it's target and format and bioavailability.

25:19Peter Attia:Toxicity, the whole thing. Everything. And all of them have maybe a 90 % chance of success. You multiply all that out 10 or 20 times, it's zero. It's a huge failure. Yes. Yeah. So you have to minimize risks at every step other than the ones that you sort of identify and accept. This is the big question that we don't know and we have to answer early. So we want to - And you want to fail fast. If you can. Yes. Yeah, absolutely. The worst outcome in drug development is failing in phase three. Oh, yeah. Actually, that's probably not true. The worst outcome is succeeding in phase three and failing commercially.

25:57But failing early is super important.

26:00Peter Attia:Now, let me just ask a question about this specific indication. So you've already made a decision, which is we're going after sarcopenia. And you've decided to do that as opposed to say, we're going to go after muscular dystrophy. So are you doing that because sarcopenia is a much, much, much bigger market, which is the obvious choice? Or are you saying it's easier to get approval there and then ultimately we can also demonstrate that this will work in Duchesne's muscular dystrophy. How would you think about those two ways to proceed? So my personal bias is I have a limited amount of time on this earth to develop medicines and help people.

26:42I want to help as many people as I can. So I mostly do large indications. And that's my personal choice. Now, in that situation, regulatory pathway might

26:53Peter Attia:be harder for sarcopenia if it is not yet identified clearly as a disease. No question. No question. Whereas Duchenne's muscular dystrophy, that's an orphan disease, you would probably get a quicker regulatory pathway to approval. Yes. Yes. And so a different strategy might be we go after Duchenne's because we can get there faster. And then once we have demonstrated a drug that works in that category, we then chase the approval pathway to help as many people as possible. Would those just be two different strategies? Yes. And frankly, I've done both depending on the circumstance and the mechanism of the drug and so forth.

27:29Peter Attia:And frankly, I'm doing that right now with one of the two companies that I work with. Okay. So let's go back to your example though. Let's get back to sarcopenia. The very first thing that I tried to do in sarcopenia was prevent falls. Now, So how do you measure a fall, Peter? Well, I was going to ask a harder question, which is, which are the muscles that are most responsible for a fall? And maybe that's part of what you could test. It turns out, I think it's very complicated, right? I've had a guest on this podcast who presented data that suggested something that's seemingly as innocuous as great toe strength is an enormous predictor of falls.

28:11Peter Attia:Now, there's a very objective way to measure the force that a person can generate with their great toe. And as that force gets below certain thresholds relative to their body weight, the probability of falls just starts to go straight up. Balance, obviously. I would argue that they don't really know that. Because I asked the question, what causes falls? There's about 11 different things that cause falls. And people can think through themselves that are the obvious ones like weakness and dizziness, which dizziness itself is very complicated, but also vision, also attention, lots of things can cause falls.

28:49Yeah.

28:50Peter Attia:I mean, I think another big one that we see clinically, Lloyd, is loss of reactivity. So foot speed and reactivity. So you and I, if we went for a walk today around Lake Austin, so there's a 10 mile beautiful loop around Lake Austin, the probability that on that 10 mile loop, you and I wouldn't stumble once, at least once, and miss our footing is zero. But I would venture that neither of us would fall. And so the question is why? Why wouldn't we fall despite being challenged at a great level? You know, stepping on a twig, missing a branch or a root or something like that. And the reason is we have the reactive speed of our feet to catch ourselves.

29:32Peter Attia:And that to me is one of the things that's missing. And that tends to come down to the type 2A muscle fiber, right? Like that's a very explosive. But also proprioception. Yes, absolutely. Vision, all the things that we're talking about. But maybe the broader point is this is multifaceted. There is an atrophy that is beginning of various systems in the body, and it's creating this perfect storm where when you watch an elderly person fall, you realize that that's a situation where they would have saved that fall. It's not the insult that's the problem, whatever caused the perturbation in the step.

30:06Peter Attia:It's the inability to catch it. I think that's exactly right. So let's get back to the issue of measuring falls. But that's hard. Yeah, but this becomes a very hard problem. So I tried to do that because in a clinical environment, the only falls that are ascertained are those that cause injury. Right? From our perspective as physicians, those are the patients we see. And patients are reticent about reporting falls because they know they could potentially be taken out of their home environment if they were felt to be unsafe. And frankly, I would do exactly the same thing if I were in that circumstance.

30:47I want to stay home.

30:49Peter Attia:Yeah. So we developed a study to try to measure falls. And so we worked with a large company that manufactures triaxial accelerometers. And we made a research device for this effort. And I designed a wonderful study. So the study was we were first going to put the device on bad ice skaters in Boston in the winter and videotape the rink. And so the videotape results are the positive controls for actual falls. And then we would look at the device telemetry and look at the sensitivity and specificity of the device. It's worn on the ankle, the wrist, where we do it. This was going to be a pendant.

31:35Peter Attia:Okay. And then the second part of the study is if the first part worked and the device worked on falls that were real, we were going to put it on elderly nursing home residents. And there the positive control was little old lady found on the floor because the people were old and frail and unable to get up themselves. So when the nurses or the staff found them on the floor, that was a fall. However, they got there. And then we look at the device telemetry and see. And sorry, was the purpose of that exercise, Lloyd, to see if the ice skating telemetry could predict a fall on ice? And was it offering the same insight that you were seeing in the actual field with the elderly people?

32:15Well, it's a little bit that falls are like the Supreme Court in pornography. You know, you say you know it when you see it. And so we wanted to see it on the ice skating rink. and then look at the device telemetry and see, was it reporting falls when they actually happened? Was it reporting falls when they didn't? Was it missing things? Basically to see if the device worked. So I put that through the institutional processes for funding, and I was told, cut the first part out. And so just put it on the older adults. So I worked with a really good geriatrician named Lou Lipsitz, and we did that study.

32:52And Kieran Dole was the clinical operations person. And it was really a good study execution because working with patients and research subjects, participants in their 80s up to a little over 100 is challenging.

33:09Peter Attia:And these were all individuals in one living environment or across multiple? Yes, they were in one living environment. We wanted to minimize the variables. How many subjects were? We had 60 subjects. And you falled them for how many months? We falled them for six months. And these are people who had fallen at least once in the prior six months. So we knew they're at a high risk for more falls. And remember, this institution has a lot of protocols and procedures in place to try to prevent falls. And they're still falling now and then. So Lloyd, 60 subjects followed by six months. How many falls did you capture?

33:46Well, we ended up 117, I think, was the number of events that actually happened based on someone found on the floor. However, the device was awful in that I think it had about, it detected 17 % of the real falls. And only 17 % of the device actuations where it said somebody fell were falls. So it was just not useful. and it was the best we could come up with.

34:16Peter Attia:And just to be clear, you weren't asking it to predict antecedent movement pattern of fall. No, no. You were just asking, once a person has fallen, do you know? Yes, very simple. I would have guessed that you would have been higher. So did I, but we were wrong. This study is actually published now, but then we couldn't use it for measuring falls. Yeah. We tried one more thing. There is a Massachusetts Institute of Technology professor named Dina Khatabi who was using Wi-Fi type devices to measure people's movements in their homes. And it is a little scary. It could measure wherever you were and whatever you were doing.

35:00And so we thought that would be a great way to assess falls. But ultimately, I don't remember the issue, but we ended up not being able to use it. So we had to stop our whole program from making drugs to prevent falls because we couldn't measure them.

35:15Peter Attia:Now, let me play devil's advocate for a moment. Couldn't you have just said, we want everyone involved in a study to report if they fell because it's going to help us get better data for the study? And I mean, given that you're going to capture all the reported falls and really all you're trying to do is capture the signal of unreported falls. But if you ask the people, say, look, we're not taking away your driver's license, you know, whatever it is that you're afraid of losing, we are just trying to report this as an AE or as an outcome. I mean, it seems to me you'd have a much higher chance of capturing it than any sort of device at this point.

35:56That would be true. We were just worried about the data quality. Because people, you know, remember, these are older adults who were frail. and we were worried about the recall bias and we were worried about whether they would have something to write it down with and people generally weren't device savvy at the time. I think that might be better now. Anyway, we elected not to do it.

36:20Peter Attia:I have to think there's some AI way to do it. Like there must be, I mean, were these accelerometers using any AI? This was a long time ago. This was 10 years ago or more. So there was no AI in those days. Well, some people would take issue with that. Yeah, yeah, but not the way we see it today. Not the way we see it today. Anyway, it was an interesting example of a lot of ideas are conceptualized, and we have to be able to objectively measure things ideally in drug development. And we tried, and we couldn't, and we moved on. So this kind of brings us to sarcopenia and muscle mass and strength, which was the genesis of Bimagrimab and active and receptor antagonists in general.

37:01Now, we knew at the time about myostatin. So Seijin Lee is sort of the father of myostatin. He discovered the biology. And in rodents, myostatin is kind of amazing. You can turn a mouse into an Arnold Schwarzenegger mouse by blocking myostatin.

37:25Peter Attia:This was in the mid-90s, right? I think so, yeah. Yeah. I mean, I remember, this was when I was in medical school, I remember in 97 seeing the images of the mice, the chickens, the dogs, the cows. I mean, we, we couldn't get enough of these myostatin knockout animals. We thought it was the greatest thing we'd ever seen as students. There were no people who had that though. No. So at the time, the people who led the discovery project, so this is the laboratory research, were Chris Liu and his team and David Glass and his team. And that's where the original code of the Magromab was BYM338. And that's where the VAT came from.

38:11I was part of that team on the clinical side. And if they made the drug, how would we test it?

38:20Peter Attia:Do you want to explain how myostatin inhibition would lead to enormous muscles? Yes. So the broader question is what governs the size of your muscles? And we know it's nutrition and we know it's use. And then there are biochemical things that regulate it. And myostatin is an inhibitor of muscle growth. So if you inhibit the inhibitor or block myostatin, muscles get larger up to another point where there's something that regulates them, and we don't know what that is. And that's most of the inhibitory effect biochemically in animal species. In humans, it turns out that it's more complex. It's myostatin plus activins, mostly activina.

39:09And this is the advantage of inhibiting myostatin and activin A together or blocking the receptor, which is what the Magrumab does.

39:22Peter Attia:What was the evolutionary, I mean, I'm asking you as though you were there during the design phase, but what is your best guess for what the evolutionary reason was to limit muscle growth? Was it simply a nutrient management system, which was like, hey, we grew up in a nutrient sparse environment. We can't just have muscles demanding all of this protein. I would have to guess that based on everything else about people, there were times during our evolution when resources were very scarce and we had to conserve. But I don't know. Your guess is as good as well. But it is interesting because the other way that nature could have solved that problem was just to make it completely supply limited and just say, yeah.

40:03Peter Attia:But anyway, okay, so we don't really have a great teleologic reason. That's kind of the way it works with adipose tissue. Yeah. Yeah, exactly. That's right. So we don't have a great explanation for why, but regardless, in humans, myostatin plays a smaller role than it does in these less, presumably slightly less complex mammals. And what is the actual mechanism by which myostatin is inhibiting? Is it doing something in actin myosin filaments? What is it doing to prevent hypertrophy. Complicated. But to summarize it briefly, the receptors are part of this larger TGF-beta superfamily of receptors, and there's dozens of them.

40:44There's type 1, type 2, type 3s, and they all signal via mostly a common pathway called SMADS, which was named by those whimsical drosophila geneticists. It stands for similar to mothers against decapentaplegic, which people don't need to know. And those are transcription factors, and they govern a whole lot of gene programs. Some of the more important ones are muscle. So muscle size is regulated by nutritional availability and then muscle protein synthesis versus muscle protein turnover. And the proteins that turn over muscle are MRF1 and MAFBOX or atrogen. And David Glass was one of the discoverers of this pathway, who I mentioned earlier.

41:41And myostatin signaling via the active in receptors suppresses the proteins that are involved in targeting muscle proteins for degradation.

41:55Peter Attia:So before we leave myostatin to talk about BEMA, do you want to say anything about folistatin? And I don't even know if you're aware, but folistatin became a very popular recreational sort of gray market agent that was sold for research purposes only in quotes. and the marketing material suggested, look, if you take folistatin, folistatin inhibits myostatin, you're going to get really big muscles. And so people were pumping themselves full of folistatin. Actually, that's not really true. I think the folistatin was so expensive. They weren't doing that. They were doing some attempt at folistatin gene therapy.

42:33Peter Attia:So do you want to just explain what folistatin is or why it may not be as, uh, as, as Holy grail as it was made out to be? So folistatin and then there's folistatin-like proteins are endogenous inhibitors of this pathway we've been talking about. And it does, if you do a gene therapy in rodents, result in larger muscles. But it's a small protein with a relatively short half-life, and I don't think dosing it systemically now and then would be successful. We did the math on this, and you would need to give it several times a day. and given the price of folistatin, you'd be spending above, I don't know, you'd be spending a million dollars a month on folistatin.

43:18Peter Attia:But also it wasn't clear that it would do anything in an adult. In other words, it seemed that there might've been a critical window during which administration of folistatin or folistatin gene therapy would have an impact, but it had to be pretty young. It'd be during the development of the muscle more so than a mature phase of the muscle. I think it would work in adults if you could solve the half-life. And I know there are companies working on this with FC fusion proteins and other half-life extended versions. That's a good point. Can you tell people why an FC fusion pro... I mean, I think we have to talk about some of this technical stuff, unfortunately.

43:53Peter Attia:I almost caught myself as I was asking the question, but explain what an FC fusion protein is and why that might be able to keep it in place longer. Sure. You can edit this out later if it gets too technical. I'll try to speak in plain English. Well, yeah. Just think about it through the lens of how you manipulate drugs. I think of this as part of the story. Right. So if I remember correctly, the technology that we use now for FC fusion proteins was originally developed by Brian Seed at the Mass General. And the very first drug that used it was Enbrel, which is etanercept, which is itself a really interesting story because This is a TNF inhibitor that's now approved in rheumatoid arthritis and psoriasis and some other things.

44:39But it was first tested in sepsis. I didn't know that. And it made people worse. Well, because I think physicians and scientists at the time knew that sepsis was an exuberant inflammatory reaction to an infectious stimulus. And they thought by tamping down the inflammatory component, you might be able to have better outcomes. And it turned out it made people worse. but ultimately it was then tested in rheumatoid arthritis and it was amazing and you know the story after that but a company called immunex at the time licensed the technology from mass general to make a tannercept and essentially what it does it does two things so the soluble receptor which was the which was the low affinity regulatory tnf receptor i think at the time I've got to double-check that.

45:28I haven't thought about a tinercept in like 20 years. Had a relatively short half-life just as a protein injected all by itself. And doing the recombinant DNA technology to attach it to this part of an antibody called the FC region did two things. One is it extended the half-life of the protein in the circulation by allowing it to recirculate the way some blood proteins are recirculated normally. So you're sort of hijacking an endogenous mechanism to preserve proteins in the bloodstream and doing it the same way. And the second thing it did is it put a hook on the protein to allow you to purify it easily because the protein G and protein A columns were well-established at the time and easy to use to purify proteins.

46:22So that's what it did for the drug developers. And since then, the technology, this is a great example of going back to our patent discussion. Originally, that was patented technology and only ImmuneX could use it or some other licensee of Mass General, but the patents expired and now it's, the technology's freely available to the rest of the world for the rest of the eternity. And many, many, many companies use this technology to make drugs and we're all better off for it. Okay.

46:53Peter Attia:So how did you guys discover BEMA? How did you create it? So again, the earliest biology was done by Chris Liu's team in the Pathways group at Novartis Institutes. Jeff Porter was the leader of that group. And the idea was we wanted to inhibit the receptors, not go after all the possible ligands, because we knew that myostatin wasn't the whole story in humans. We didn't know which activins it was, thought it was probably activin A, but it could have been others too. And that story and that thinking has panned out subsequently, as we can get into later. And at the time, therapeutic antibodies were the best technology to do this.

47:40Remember, the affinity of the ligands, myostatin and activins for the receptors was nanomolar, low nanomolar, maybe high picomolar. So we needed an inhibitor that could bind down in the low picomolar range in order to effectively prevent ligands from binding, can't really do that easily with small molecules.

48:05Peter Attia:Okay, this is a great example of the importance of understanding that distinction. Yes. I'm going to explain what you just said, and then I want to have you state that last point again. So people, when they're talking about pharmacokinetics, they're talking about affinity. You'd have to talk about a concentration. So what concentration of this hormone or this ligand is necessary to get into this? And when you start talking about, you know, we can talk about millimoles, micromoles, nanomoles, as we get smaller and smaller and smaller, as those numbers get smaller and smaller and smaller, it means you don't need very much of the thing to get in the receptor.

48:43Peter Attia:And therefore, if you're trying to develop something to block that, it becomes a harder problem because you better figure out a way to usurp this guy getting in and this guy gets in very easily. Yes, exactly. Exactly. And you're saying to get a small molecule to have that degree of sensitivity is very challenging. Yes. And therefore something that's biologic makes more sense. Yes. And is that due to just the physics of the conformational fit? Yeah. I think you could describe it that way. It's essentially this, it's more complicated than this, but think about it as the surface of interaction of the two molecules binding together.

49:26Essentially, we need, if the ligand, myostatin and activin for the receptor is sticky, you need an inhibitor that's even stickier. And that would be very challenging to do with small molecules. So in this case, we went with the biologic route, sort of ran a therapeutic antibody project in collaboration with Morphosis, with whom we had a collaboration at the time, and had a bunch of candidate antibodies and then ran through the usual developability, maturation, improvement of what we get in the screen until we have what we think could be a therapeutic drug.

50:03Peter Attia:How many molecules enter the top of that funnel? Oh, there might be thousands of antibodies that get tested initially. This is done with something called phage display technology. So I would say for the past 15 to 20 years, we don't make antibodies in mice anymore. It's all done by recombinant DNA. DNA using these viruses. Okay. And I think back, I've been in the business long enough. You know, I've made antibodies by immunizing mice and fusing cells and growing them up and putting them back into mice as a CITES tumors to get enough antibody to do experiments with. I mean, it was terrible. It's much better now.

50:46Peter Attia:So you're literally running a screen with all of these antibodies and you're screening for two things, but well, basically one thing, what is going to give me the lowest concentration that binds to this ligand, that binds to, sorry, to this receptor? Yes. So knowing that you have to be below a certain concentration. We knew we needed really high affinity antibodies. There were some things we didn't know. So in the laboratory, most of the time you can measure a cell's surface receptor on the surface of the cell pretty easily using things like flow cytometry. The active and type 2 receptors are actually expressed at such low levels, you can't see them by flow cytometry unless you somehow very artificially manipulate the cell.

51:31So we had to create a screening assay that was essentially a reporter assay. So we couldn't measure the receptor on the surface of the cell. We could have developed an assay to do that, but it would have been very laborious, radioactivity, not necessary. So we put in a reporter gene, which is, and basically made the cells glow with firefly luciferase if the ligand bound myostatin or activin. And then we were looking for decrease in the glowing with therapeutic intervention. And we also didn't know whether we needed to inhibit activin receptor type 2A or B or both. Most of the work in vitro suggested for muscle hypertrophy, most of it was driven by 2B, but we weren't sure.

52:23And we ended up getting a 2B preferential drug, but it also hits 2A.

52:31Peter Attia:And this has nothing to do with muscle fiber type? No, this works on all fiber types. Oh, okay. And then we looked for antibodies that worked in that cellular assay where we wanted to prevent cells from glowing when we added myostatin and activin. It had to work on both of them. In other words, it shouldn't matter which ligand you put in. The antibodies should block their activity. And the affinity of the antibodies had to be really good. So it had to be much stickier than the ligands for the receptor. And so we had that. And then the real important experiment is could we block the activity in an animal?

53:11Peter Attia:But just that first step, Lloyd, until you could identify candidates, how many months was that? Years. That was years. So again, just going back to the analogy of building a skyscraper, that's the planning phase of the skyscraper. That's the excavation of the hole. That's probably the laying the foundation. You haven't actually put any of the big pillars up yet. It's the permitting. It's securing the funding. It's sort of the site planning. It's all of that from your building perspective. And by the way, I assume that the standard estimate 20 years ago was every approved drug is approximately 10 years and a billion dollars.

53:52Peter Attia:That's got to be pretty low today. Do you have a sense of what the more accurate dollar figure is? It's got to be more than a billion today per approved drug. Oh, yeah. There's a Tufts organization for the study of drug development, and they think it's, I don't know, two, three, four billion, something like that. Yeah. The part you're doing now is time consuming. Luckily, it's not that expensive, correct? You're spending millions of dollars, but not necessarily - Exactly. You know, tens of millions at this point. Yes. Yes. So you finally identify a candidate or several candidates that you now want to take to the next step, which is, hey, in vivo, does this thing work?

54:30Yeah, so we would typically have two to ten at this stage. And it's taken years because we have to work through the biology and make sure that we understand what's going to happen. Not because we want to save the cells in the Petri dish, but because we don't want to start something unless we have confidence it'll succeed if we pass each step. So we have to work through the biology. We have to make all the tools we need, which are these glowing cells in response to myostatin, for example. And there's plenty of other tools that we need, too. We need to make the reagents. We need to make the myostatin and the active NA and all the tools we need to do these experiments.

55:14So, and then run the experiments. And we have to make the antibodies. And that takes quite a bit of time as well as developability on the antibodies, which means we need to, at the very earliest stage, have high confidence that we're going to make an antibody that we could give to people and it'll be stable. And it'll have predictable physical properties and it'll have a shelf life. And all of these things are what we build into the antibodies at this very early stage.

55:44Peter Attia:How confident are you at that stage, Lloyd, that the antibody won't elicit an immune response in a human? It's still one of the biggest unknowns. And the best way to do it is to use fully human antibodies. Not humanized, fully human. Straight up human. Yeah. Can you explain to folks the distinction there? Yeah, well, humanized is used to describe taking a mouse or other species antibody and replacing as much of it as you can with human sequences based on what we know about human antibody codon usage and amino acid preference and so forth. Whereas fully human means you're starting with human genetic material antibodies.

56:28Peter Attia:But ultimately, it still has something in it that's foreign. Thank you. As little as possible. Part of the sequence. I mean. As little as possible. There's going to be something new because antibodies have this intrinsic ability to recombine and create new sequences. Yeah. Now in vivo, in people, so if you make some brand new antibody and it's not suitable for some reason, it gets selected out. Whereas when we do this in vitro in a test tube, that doesn't happen. We do the best we can and try to end up with sort of common codon usages, common antibody sequences, even pairwise and so forth. But you don't really know until you put it into people.

57:12There are some in silico screens you can do by looking at what peptides are likely to be generated in a lysosome, and are they going to have high affinity binding to an MHC molecule and so forth, and we do all that stuff. But you still don't know until you do it. Now, it's an interesting historical perspective, because I'm getting old enough to be interested in history now. The very first antibodies tested in humans were mouse antibodies. And there's still one that's used to this day as a therapeutic. So it's OKT3. So this is an antibody used to prevent transplant rejection. It's an anti-T cell antibody, human T cell antibody.

57:52And we still use it to this day.

57:55Peter Attia:And what is it targeting? Is it literally targeting CD3? Yep. Wow. So it's broad. Yes. It's going after every T cell. Yes. Right. And there's rabbit antithymocyte globulin, too, that's still used clinically. That's a blast from the past. It's still used. Yeah. And these were the first antibodies used in people. And of course, the other foreign antibodies that are still used are antivenoms. Most of them are horse serum. As you can imagine, you only want to get bitten by a snake once and need that. Because when you need it the second time, you have kind of a ferocious serum sickness response. Great point.

58:36Peter Attia:So you identify BEMA plus a few others and you start now running these into the mice. Yes. Is there another chance that because you've gone to all the trouble to make sure you have a human antibody, it won't work in the mice when it otherwise would have worked in humans? We test that beforehand. Okay. Got it. So this is, again, there's an enormous amount of detail in drug development. And one of the things is that your drug has to cross-react, I mean, work also, in at least one of the two species we're going to use later for toxicology. And it has to work in a species we're going to use for pharmacology.

59:22If it doesn't, then you need to make surrogate drugs to do that. But in our case, we made one that worked in rodents. Now, it turns out that bimagromab is pretty immunogenic in mice, but not in rats. Although, how common is that? It's kind of idiosyncratic. So the construct we ended up using in mouse experiments was something called CDD866, which was, you talked about humanizing antibodies. we murinized the Magromab so that we could use it in mice.

1:00:00Peter Attia:So you could go back and use it. Yeah. That makes sense. And to make a long story short, we were able to give these antibodies to mice and see if it caused muscle hypertrophy or not. And how much did it do so relative to the pure myostatin knockouts that were enormous? Probably more so than the myostatin knockouts. Wow. It was really impressive. again we're going to link in the show notes to what these images look like but it is it is truly a caricature it's impressive in the mice look at also the whippet dogs the the belgian blue cattle that are double muscled it's uh it's impressive did you do anything else lloyd at that time so when you demonstrate that bima is making bodybuilding mice was there any assessment of muscle function?

1:00:51Yes.

1:00:52Peter Attia:Okay. And what did you find? The mice were stronger and can run faster. Okay. But remember, they had perhaps a 30 % increase in their muscle mass. I mean, which is huge. And people can look at the pictures online and see it's quite obvious. So it doesn't work this well in humans. Just skipping all the way ahead, and we'll come back, humans get about 4 % to 8 % increase in muscle mass. Most of the people we've tested have been older people, which is one caveat, whereas... Did you do it in old mice? Not in old mice. I think David Glass did some experiments in older rats, and it still worked, but not as well as in younger ones.

1:01:39And remember, when people do these muscle experiments in rodents, they almost always do males because it works better in males than females.

1:01:47Peter Attia:Fool thyself. Um, it would be interesting to know as you ran it across a continuum of escalating age, what the, what, what accounts for the reduction in efficacy, right? Is it, is it substrate limited? Is it a muscle protein synthesis problem? I mean, what, what is, you know, I don't think, I don't, I don't think that's been studied formally by scientists. It might be known in the, in the bodybuilding community. And just a heads up to the listeners who might be bodybuilders, the first thing that any company does when they're working on a drug that has the potential for abuse is we work with WADA, the World Anti-Doping Agency, to make sure that they can screen for these things.

1:02:32Peter Attia:So where in the pathway of BEMA did you begin notifying WADA that we're working on this thing. As soon as we had a therapeutic antibody, we started that process. And basically, they've had an assay for more than 10 years. Have they ever caught it? Have they ever screamed? I have no idea. That's funny. Okay. So after you get the home run in mice, do you want to go into a primate? Where do you typically go from mice? So for therapeutic antibodies, often it includes a primate simply because we want to have one of the two toxicology species in whom the antibody has the expected pharmacology so we can look at sort of on-pathway and off-pathway toxicity of the therapeutic.

1:03:27So we typically use non-human primates for this.

1:03:32Peter Attia:When you're at the mouse stage, Lloyd, how are you screening for tox besides the most obvious, right? Obviously, mortality or something catastrophic is obvious. But for non-apparent or non-mortality-based toxicity, what are you looking for? And is any toxicity at the mouse level disqualifying to go forward? Or are you evaluating it case by case and saying, look, okay, this ended up being pretty bad for the mice despite the efficacy. We don't think that's going to be an issue or we think we got the dose wrong? Or do you basically keep going back and perfecting it in the mice until you get the dose response right before you move up?

1:04:06Peter Attia:Or do you just sometimes say, no, we're going to go to the primate or whatever other model we're going to look at and reassess tox as we get closer to our species of interest? Yeah. So with the caveat that I'm not a toxicologist, the fundamental principles are you You use a weight of evidence approach based on all the data that accumulates. From a clinical perspective, we try to balance risk and benefit with new medicines in general. So if it turned out that bimagrimab had some, and bimagrimab has some talks that we'll get into, but if it were unsuitable for use in a big population, that we then think about higher medical need patients.

1:04:49Peter Attia:That's when you would go from maybe sarcopenia to Duchenne's muscular dystrophy. Yes. And where whatever the adverse effects are would be outmatched by the potential benefits. Yep. But secondly, we want to know precisely what the toxicology is. And the two big things we look for are whether it's monitorable, whether it's reversible. So if we have irreversible cardiac toxicity with a therapeutic, that's usually the end, for example, or neurologic. If it's serious organ toxicity. But there's enough of a prodrome. So for example, if it's, well, you look at a drug like Lamisil, right? Something as supposedly benign as Lamisil.

1:05:38Peter Attia:I mean, that can destroy your liver, but there you can stop the drug. So that's your point. It's moderatable and reversible if you stop it and you get a long enough warning. Because otherwise, if it would just automatically destroy a person's liver at a frequency of one in a hundred people using it, you can never justify it. So you're describing idiosyncratic liver toxicity, which is the most common reason drugs get pulled off the market still. And I personally have killed drug programs for that. And it's hard to, can't predict it preclinically. Yeah. Yeah. So, so let's get back. You get BEMA into the primates.

1:06:19Peter Attia:Yes. And how does it, how does it perform? Well, in order to do toxicology studies, you have to know how much to give them and how long it's going to last and what it's doing. So there are, there are preliminary studies in a small number of animals. And so we did those, but we did them long enough so that we could see the muscle hypertrophy of what were going to happen. And it did work, not as well as in the rodents, but it did work. So it gave us confidence that we could move ahead with the rest of the activities. Now, manufacturing enough antibodies for use in larger animals is time and expense.

1:06:59And so all of that was going on in parallel.

1:07:01Peter Attia:And each of these decisions, so you're doing this all inside of Novartis. Is there an, I see, is there an investment committee that basically revisits every time there is a new, you know, allocation of capital to move from one thing to the other, where everybody presents and how does that typically work in a large company? All big drug companies work kind of the same way in that there are, there are typically two or three or four, depending on the company, major checkpoints where all the data are assembled and, and made into a slide decks and presented and feedbacks obtained and programs, courses adjusted.

1:07:38And so that happens. And I think it's more frequent, but faster at small companies. But things are constantly being reevaluated and reassessed.

1:07:49Peter Attia:Yeah. And what's interesting for folks listening to us is we're talking about this in the context of a large company that doesn't have to go out and raise capital every time it does this, but the exact same idea that you just described, everything you just said could have been done by a startup, but now it would have a totally different look and feel in that, hey, we're going to go raise some seed funding to go test this idea. Okay, guess what? We were able to find the antibody. We're going to have to go raise another, you know,$20 million and boom, boom. And now they'd be at the stage where they'd be probably raising a series B or no, probably this would be a, still be an A, I think, as they go into the primate.

1:08:26Whatever you call it, you typically -

1:08:28Peter Attia:But give folks a sense of how much you'd have to raise for this next stage, which is basically your pre-IND. So if it includes many, so you have to manufacture - Yes, it's going to include the primate through manufacturing. Toxicology and clinical material. You have to run the IND enabling studies, which is toxicology and some other things. and frankly, investors want value creation for their money and reasonably so. So I would think for Bimagramab, if this were in a small company, the value creation step would be showing muscle hypertrophy in the very first clinical study. So the funding that I would raise would be IND enabling plus phase one, plus phase one, plus a runway to raise the next round.

1:09:19Peter Attia:Yeah. And you would structure your phase one to demonstrate efficacy, even though technically you only need to do talks. Yes. You would have it long enough, big enough. Not talks, but safety and tolerability. Yeah. So, okay. And then just for using BEMA as an example, how many dollars would that be from where we are now to give you that runway into 2A? In today's dollars, probably$20 million. Okay. Yeah. So Series A. Thereabout. Yeah. Okay. Maybe a little. That's lower than I would have guessed, by the way. Maybe a little more. Okay. It depends on where you do the manufacturing. Yeah. So for Novartis, this is nothing.

1:09:57Peter Attia:For a startup, this is everything. They're betting the farm. Well, not enough. I don't want to make light of it inside of Novartis, but the point is Novartis doesn't have to go back to the public market to say, I need to raise another $25 million to fund this. They're doing a lot of these in parallel. Big companies though have have, their resources are stretched too. It's kind of funny thinking about it, looking from the outside, but having been inside a big company, people are competing for a fixed amount of research dollars, the people within the company, and resources are allocated based on company strategy.

1:10:30And ironically, sometimes there's more project capital available in a small company than in a big company, because the small company's got one or two or three projects, all the money's going there. Whereas there are hundreds in the big companies. And I've seen it both ways. I've seen some big company projects get high profile, high importance, well-funded. So yeah, so we'll say$20 to$30 million maybe at this stage. and the Magromab at that point made it through rodent and non-rodent toxicology studies and what we call DMPK, which is distribution metabolism pharmacokinetics. It's knowing that when you give a participant or a patient, a subject, a medicine, does it get into their body?

1:11:27Does it go to where you want it to be? Does it do what you expect it to do. You have to know all that stuff before you go into patients for the first time. And we assess that in animals. You asked earlier, what do you actually do to measure the toxic effects of a medicine? And animals receive courses of therapy and we do blood tests just like we do in people. Sometimes we would do x-rays if it was warranted. And then they get autopsied to look at all the organs and look for microscopic changes that you might not perceive clinically. And we have to know all of that before we give people an experimental medicine for the first time.

1:12:09So any red flags

1:12:12Peter Attia:whatsoever as you, or anything that is of concern, not necessarily a red flag, but anything that's still an unknown as you're going into the phase one? Lots, lots of unknowns going in. And there are always things of concern. I've never seen a drug development program that couldn't be stopped for some reason. And you have to balance the unknowns and the uncertainties and the risks with the potential benefits and make a decision about whether you move forward or not. It's kind of a joke in the industry that every really successful program has been almost killed or killed several times before it eventually makes it out into humans and then eventually commercialization.

1:13:01Peter Attia:What's the approximate attrition from that first candidate drug discovery to the IND filing? That's a winnowing down of what to one. It's hard to put in an aggregate because it depends based on the format of the drug. And then there's other factors like strategy and funding and everything else. But for biologics like bimigrimab and a therapeutic antibody, it's pretty low, actually. Five to one? Six to one? I would say maybe 30 % of them actually could get into humans. Okay. Yeah, more than I would have thought. Yeah. It's simply because there are no off-target adverse effects with antibodies in general.

1:13:50There are some specific counterexamples to that, but in general, an antibody is not like a small molecule that could have liver tox or some other tox that you can't predict for reasons that you don't understand.

1:14:03Peter Attia:That's a great point. Yeah. So maybe I'll restate that so folks get it because, because the antibody is so specific by definition, it can't bind to many other things. And in fact, we screen to make sure it doesn't. Yeah. Yeah. Whereas the chemical can do lots of things off target. You know, I had on recently, we had a podcast talking about C-TEP inhibition. And, you know, the very first version of that drug lowered LDL cholesterol, but raised blood pressure. And that was a completely off target complication of the drug. Yeah. Exactly. So that doesn't generally happen with biological. Got it. So that's why you have the higher throughput.

1:14:42Peter Attia:Yeah. Okay. Yeah. Let's skip. So you file the IND and you're now ready to start a phase one. So an IND is requesting regulatory permission to administer the drug to people. And you've already filed your patent at this point? Yes. Yeah. Where in that process did you file it? Patents typically get filed. Again, there's a... You want to do it as late as possible, but while still protecting. Exactly. So typically around the point where you have a group of candidates from which your final drug will be selected, that's typically when we would do it. Okay, wow. Because you want the patent to last as long as possible, but once information about what you're doing is getting out, you want to have it protected.

1:15:31Peter Attia:Okay. So from the time you file the IND with the FDA until, and you have to show them everything that we've talked about. Do you also have to, at the IND, show them that you can manufacture in GMP? Yes. So the manufacturing is a core element of the OMN application that you do for an IND. And this is U.S. specific nomenclature. IND stands for Investigational New Drug. In Europe, it's called a clinical trial application, CTA. There are other countries that have different nomenclature, and companies can do the first in human study anywhere in the world that's got a proper regulatory environment and suitable investigators and clinical sites with adequate quality and so forth.

1:16:21But maybe we'll be U.S.-centric for this discussion.

1:16:26Peter Attia:Can you explain to folks what the hurdle is to GMP or good manufacturing processes and why it's so important? And again, I call this out to listeners because we live in an era now where these peptide therapeutics are very prevalent, these sort of gray market peptides. And there are people out there that think, hey, I'm buying retitrutide. Yeah, no, they're not. Yeah, exactly. Maybe use the GMP process as a way to explain why when you think you're buying retitrutide peptide for research purposes only, you are definitely not buying what Eli Lilly is going to eventually sell if they get FDA approval.

1:17:07Right. So GMP stands for Good Manufacturing process. And it's basically a commitment by the manufacturer to use high quality standards with extensive documentation to be able to prove what they've made so that everybody can have confidence that this is a good quality material and they know that what's on the label is what's in the bottle, basically.

1:17:38Peter Attia:And that there's nothing in the bottle that's not on the label. Exactly. It's purity, it's activity, it's contamination or lack thereof, it's sterility, if you will. It's all of those things. It's that the material that's being purchased eventually commercially is the same material as what was tested clinically, and we can have confidence in it. Basically, the factories are inspected and the factories that make it have, and typically there are many manufacturers involved. When you buy a, with the exception of saying buying a bottle of terzapatide from Eli Lilly, but typically when you buy a drug from somebody, the drug's substance, the chemical, is manufactured by one company And then it is formulated or put into a mixture that makes it predictably absorbed or administered, is done by another company.

1:18:39And then it's put into a package by a third company. And then it's distributed by a fourth company. So there's a lot of people involved. And this whole manufacturing infrastructure and pipeline is well-controlled and well-documented. And you can buy online peptides that may be the same as Reditrutide. They might not. You have no way of knowing. Yeah.

1:19:04Peter Attia:And in many ways, that's the premium you're paying when you're buying the drug from Novo Nordisk or Eli Lilly or Novartis or whatever is the part of the premium is it's very expensive to manufacture under GMP conditions. Yes. So it's a bit of a buyer beware when you decide not to. I think it's a big mistake to buy these peptides from fly-by-night manufacturers. Conceptually, it's no different than going in a drug user, going and buying some opioid from a street corner drug dealer. You have no idea what's in that. Could it have fentanyl in it? Could it have car fentanyl in it, which is even worse than fentanyl?

1:19:42Baking soda, you have no idea what's in there. It's the same thing with these peptides. You have no idea. I think it's a mistake. And plus, even if you were to have confidence that those peptides were what they're saying they were, the data to support what they do are almost non-existent. I've been reading in the popular literature about this one that's, I think it's called BP-197.

1:20:04Peter Attia:BPC-157. 157, yeah. All of the data for that peptide come from one investigator who's the only person published on it. And remember, the fundamental tenet of science is if it's real, it's reproducible. This has not been reproduced. What the hell is it? It's not encoded in the human genome. So it's not a human peptide. And it has no known receptor. Yeah. So we don't even know how it works. There's so many red flags for this. No, it's the poster child for what I would argue is the absolute greatest grift of the entire health and wellness industry. There's a lot of, I'm sorry if I'm insulting you, Peter.

1:20:47There was a lot of grift in the health and wellness industry. Oh, you're not insulting me.

1:20:51Peter Attia:But that's my point. Despite how much grift there is in the health and wellness industry, I'm putting BPC-157 on the podium, at least. I would, too. Yeah. I would, too. Okay. So we've made Bimagrimab, and it's made it through all of the IND-enabling study activities. And we're ready to give it to people. Who do we give it to? And this depends on what we need to measure and what the expected safety and tolerability issues are in people. Again, we talk about toxicology in animal species. We talk about safety and tolerability in humans. And again, most antibodies, including bimagrimab, won't have safety and tolerability issues that are off the pathway that it's working on.

1:21:41and didn't really see any safety and any toxicology to speak of in the animals. The only thing that I was a little worried about that we saw was in the rats, they had cardiac hypertrophy. However, remember the animals had enormous change in their body size because of the muscle hypertrophy. And if you normalize the heart size to the body size, it was normal.

1:22:09Peter Attia:So does that mean that you didn't know if the cardiac hypertrophy was in response to more resistance that the heart had to work against, or whether the antibody was working directly on the cardiac myocytes and increasing hypertrophy there as it was in the skeletal muscle or both? Exactly. We didn't know. But we could make an argument that rats of the size that they became… Should have bigger hearts. Should have bigger hearts. And that was the argument we made to regulators. And so we didn't think there was any specific cardiac toxicity. And typically in toxicology studies, you have something we call a recovery period where the drug is withdrawn and some of the animals are followed to look and see whether any toxic effects that did occur are reversible.

1:22:59And in fact, when you stop giving the animals Vimagrimab, the muscles got smaller and the heart got a little smaller. How often was it dosed? So the way you dose in the toxicology studies is you want the exposure, which means the amount of drug in the blood, to be ideally higher than what we ever expect to get in humans.

1:23:22Peter Attia:And then when you got to the humans? So that was the preamble to the answer to your question, which was weekly. So we gave the animals the drug weekly. And the half-life of the drug presumably is short, but... It's shorter in animals, but again, you drive the dosing in the toxicology studies to make the amount of drug in their blood ideally higher than we will get in people so that we have what we call a safety margin of exposure. How do you know at that point, Lloyd, if toxicology is driven by peak or trough? Because some drugs are— You don't. You don't. You don't. You make a best judgment. Is there a general rule of thumb?

1:24:02You measure both and you want both of them to be higher in the animals than what you get in people. Now, this is in general medicine therapeutic indications. In some nasty oncology drugs, toxic effects and therapeutic effects are at the same exposure or even lower sometimes.

1:24:20Peter Attia:Yeah. But the medical need is so great that you accept the toxicity. Well, I was going to actually use that as an example. We sort of skipped ahead a little bit on the phase one patient selection. We rushed through that. But, or actually we, we didn't answer it. We got, we, we, we went off topic. We're going to come back to which patients do you select for the phase one? And that really depends on the drug because in an oncology drug, you're going to test it on the most recalcitrant cancer patient, right? You're going to test it on a patient who's progressed through every therapeutic. They have stage four version of whatever cancer you're testing.

1:24:54Peter Attia:And this is the Hail Mary. And you're not just testing for tolerance and side effects. you're hoping to get a sliver of efficacy through dose escalation. That's the most common scenario in Kassar. But here, what are you doing? Are you going out to the most frail, sarcopenic, elderly person, or are you going to test it in? So what I do personally, and a very experienced drug developer named Bob Schmouter taught me this, and I think he's right, is ideally you'd like to test this, a new medicine, in the cleanest population you possibly can, where anything you measure is related to the drug and not some underlying disease or other thing.

1:25:38However, we do not want to expose healthy volunteers to risks if we possibly can. So we use our best clinical judgment to say that I don't want to expose people to a risk greater than that of a lightning strike in a year.

1:25:58Peter Attia:Is that literally a probabilistic formula you use? That's what I use. Interesting. Yeah. So the risk of being struck by lightning in the U.S. in a year is about one in 100 ,000. That's actually higher than I would have thought. Me too. That's a little scary. But that's what it is. Okay. And so I don't want the risk of something bad happening to one of my volunteers to be greater than that. And that's frankly how I explain it to him. So if I can have some confidence that that's true, we will test drugs in healthy volunteers. If we're worried about a toxicity or a risk, then we will go into people who have a potential benefit from the therapy so you can make a risk-benefit argument.

1:26:44And this is all laid out in plain English in the consent forms.

1:26:48Peter Attia:I mean, so first of all, that's a great framework, Lloyd, which is if the risk of adverse event is greater than one in 100 ,000, we must move to a population that is going to potentially get benefits to justify. Serious adverse. Yeah. Is that a Lloydism or is that a truism across the entire industry? Is that something the FDA would ask of every company? It's a schmouterism. Bob, if you're listening, thank you. But the FDA doesn't force that? The FDA doesn't force that, but the principle is still there. Okay. But it's a great standard. Yeah. Again, and if you think about the industry as a whole, how do we do in bringing new medicines into healthy volunteer populations?

1:27:31I've been in this business partially in academia, wholly in industry, for maybe 30 years total where I've been watching this. And about once every 10 years, we see something serious happen to healthy volunteers. Once every 10 years in a study. So that's pretty good. And we learn something when those happen. So we're talking about therapeutic antibodies. The one that comes to my mind and maybe to others is the Tegeneron incident. Say more about that. I don't remember that. This was a therapeutic antibody that was directed against CD28. It was an agonist antibody. Now, CD28 is an inhibitory receptor on T cells, and the idea was, I'm sorry, is an activating receptor on T cells.

1:28:16And I forget the actual therapeutic indication they were going for, but they tested the antibody preclinically. Everything was fine. And then they started at a very low dose in humans. And what we think happened is they cross-linked the CD28 receptor, and they had extremely strong T-cell activation and an acute cytokine release syndrome in healthy volunteers. Some of them died. I mean, it was terrible.

1:28:44Peter Attia:I mean, why did more than one of them die? In other words, why didn't they figure this out the very first time they administered this? That's super important. So that study, which that happened, boy, more than 20 years ago, that experience is why ever since we typically have sentinel patients and dosing cohorts when we're bringing something brand new into people. So they dosed, I think, six people at once with the active drug. Oh, my God. We don't do that anymore. Yeah, wow. The other one, there was a example with a small molecule, BiAL, I think was the example. B-I-A-L, people can look it up.

1:29:27But it's extremely uncommon to have healthy volunteers have anything bad happen to them in a drug study. Extremely uncommon. If you think of the...

1:29:36Peter Attia:I remember there was one at Hopkins when I was there. It was an... Oh, no, no, no. You know what it was? I'm sorry. That was not a, it was a woman that a healthy volunteer that underwent a bronchoscopy and I think had a horrible bronchospasm. So it was, if I'm remembering it correctly, it wasn't a drug that caused the issue, but it was a horrible adverse event to. Procedures can have. But she died. Procedures can have adverse consequences, which are known and disclosed in the consent forms. This gave me a lot of, when I was in medical school, I was, I mean, I was so broke and doing anything I could to generate a buck.

1:30:17Peter Attia:I vol, I was probably one of the most volunteered people for studies at Stanford. And like, if, if there was a study that paid a thousand dollars, it didn't matter what it asked of me, I would do it. And I, I remember coming away from that. I mean, I had radial lines in my, I had, you know, as you know what a radial line is, but arterial lines into my radial arteries that to this day, I still have scars over my wrists. Can you, can you imagine that I subjected myself to that? That seems a little much. My, I think many of us as medical students volunteered for this stuff. My personal favorite was there was, there was a study called brain electrical activity mapping that Children's Hospital was running when I was a medical student.

1:31:01and essentially they attach electrodes to your head and then you go sleep in the lab and they monitor that and video you while you're sleeping. They loved me as a subject because I was bald as a medical student and it was really easy to put electrodes on and off. I loved it because all I had to do was go in and go to sleep. But there were some others like inhaling radioactive microspheres.

1:31:24Peter Attia:I used to donate plasma via plasmapheresis as often as I could when I was at the NIH. And on one occasion I was in there and this was like a lymphocyte plasmapheresis. It's a four hour procedure. And again, it probably paid 200 bucks, which seemed like it was a lot of money. When you're a medical student, that is, that's infinite money. And then at one point, somehow the nurse stepped out and the lab locked and I was stuck in their lock and they could not find a key. So they couldn't get back in. They were losing their minds, but I didn't know it. I was just in there watching whatever movie was on the thing.

1:32:01Peter Attia:It turned out to be like one of the most stressful moments in the, in the NCI history, you know, trying to figure out a way to get a spare key to get into the lab and, and take, and I was completely oblivious to it. Oh yeah. Okay. So back to patient selection. So ultimately for BEMA, ultimately for BEMA, a healthy volunteer study. You did go with healthy volunteers. Older volunteers. Okay. So people... What was your criteria specifically in terms of muscle mass? So I actually didn't run these studies. The clinicians involved were Dan Rooks and Ronan Rubinoff at the time. But the principle here is healthy volunteers doesn't necessarily mean you're 20-something year old with no problems.

1:32:47It means people without typically diagnosable disease or concomitant medications that could confuse any assessments. In the case of BEMA, because we were thinking older adults, these were older healthy volunteers and people in whom we would be able to measure some of the effects of BEMA, we hope. And again, what you measure in a healthy volunteer study depends on what the drug is expected to do and what adverse effects you might expect. So there's some things you always do, like a set of standard blood tests. But in the case of BEMA, we were assessing people's muscle mass.

1:33:26Peter Attia:Via DEXA? And strength. This is almost archaeology at this point, thinking of what we measured, but we would have measured muscle mass. And at different times, we used MRI and we used DEXA. I don't remember what that study had. But we would have measured muscle mass who measure soluble muscle proteins in the blood like CK and aldolase and LDH and so forth. Did you see any adverse effects in the phase one? Yes. So the three adverse effects that are evident with BEMA that we think are on target and that were assessed in that study where muscle spasms or cramps. Acne is rare in older adults, and it was rare in this study because those were older adults.

1:34:14But skipping ahead, when we've tested younger people, acne is more common. We don't understand why. And then there are GI symptoms of diarrhea that happen. They tend to be first dose related and less common subsequently, but they're reproducible and we think they're real. And we saw that stuff.

1:34:32Peter Attia:And how many steps of dose escalation did you do in that study? By the way, if that's too much detail to remember, don't worry about it. I'm wondering if you remember how high you got relative to what was an efficacious dose. So there's some principles here. I personally like to dose as high as we can in the first in human study to understand if there is going to be any safety or intolerability issues in people, while at the same time never exceeding the exposures we've tested in animals. So the study designs include the opportunity to go as high as we can. Typically in antibodies, that ends up being as high as is feasible.

1:35:12And there are a lot of technical details here we don't need to get into, but when your antibodies are made from cell culture, and they're highly purified, but there's still some measurable contaminants in them. And the amount of contaminants, of course, they're also tested as part of the toxicology studies, because they're in the drug we give the animals. But we can't exceed the exposure to the contaminants either in the clinical study. So sometimes it's the volume we can administer, the mass we can administer, and so forth. So I think the highest dose that we ended up doing in Bimagramab was something like 50 to 100 milligrams per kilogram.

1:35:53But, you know, I don't remember.

1:35:55Peter Attia:And in that study in humans, you're administering once a month? Initially, you administer once, then based on the emerging results for how long that lasts. And we knew what exposures we needed to reach in order to get maximal efficacy based on the culture data. There were a lot of cell culture experiments we did that we haven't talked about. Like you can culture muscle cells in a dish, and we did that and looked at the ability of the drug to cause hypertrophy of those cells. So we knew what exposures we needed to get to and how long we wanted to do it. But again, we don't exceed the exposures that we get in animals.

1:36:34So I think after the single-dose study, we probably did three doses, and that was it for the first in human study. And typically, you need multiple doses in order to be able to see the technical term is pharmacodynamic effect, so the effects on the body that the drug causes.

1:36:53Peter Attia:So the end point for the phase one, before you move to phase two, where you're really going to actually look as your primary, and you're always looking obviously for safety, but now you're really pivoting to efficacy being the thing that you're trying to chase. Yes. What did you need to submit to the FDA to say, okay, have we checked our phase one box? Typically, you're in reasonable communication with regulators, whether it's the FDA or whether you're overseas elsewhere, and you provide them with a report. And Novartis is a European company, right? Yes. They're based in Switzerland. But this work was being done in the U.S.?

1:37:30I think we did do the first in-human study in the U.S., yeah.

1:37:33Peter Attia:Any reason for that? Are European and U.S. regulators so comparable on this point that it's really just a question of where your teams are? So every country's a little different. Europe is somewhat homogeneous, but not completely so. Every country is a little different. And I think it's true both for large companies and small companies that you go wherever makes the most sense. It's where you can remember the three biggest challenges of any clinical study are recruitment, recruitment, and recruitment. So you have to be able to get the subjects or the patients or the participants. is. You need qualified, experienced, reliable clinical investigators.

1:38:15You need a regulatory environment that's supportive for what you're trying to do. And then you think about cost of the study. They're different in different countries. And so you integrate all of that stuff. And that chooses where, at least personally, where I would go to do a first in human study. countries that are often used nowadays are Germany, Australia is pretty popular, New Zealand. Australia, New Zealand is very popular now. Things have really changed, I guess, maybe in the past year or two about China being really popular because China has a regulatory environment that's become more favorable and they can do investigator-initiated studies with less supporting data than we require for a typical IND.

1:39:05So it can often be a faster way to test something. And of course, China has a lot of patients. So that's something that's being done now too. I personally love doing studies in the U.S. and Taiwan. Taiwan has, they have wonderful investigators. They speak English better than we do. They have a very centralized clinical environment, so they have many patients at a limited number of clinical sites. The regulatory environment is very similar to the U.S. Australia and New Zealand is favorable because they have a, especially this is Australia now, they have a different regulatory construct where safety is assessed by the ethics committee and drug quality is assessed by the regulators, so they have a clinical trial notification process rather than an approval process.

1:39:59And plus the exchange rate's favorable now, so too. So if we get back to what is the trial going to cost, that's useful.

1:40:06Peter Attia:And how much reciprocity is there between agencies? So if you, well, I should clarify the question. You can conduct the trial in Australia, but under the auspices of the FDA where they're issuing, or does it have to be in the U.S. if the FDA is overseeing? If the FDA is overseeing, the study is done in the U.S. Okay. So if you do a study in Europe and get European approval, or you do a study in Australia and get Australian approval, how much of an additional hurdle is there for the FDA to typically approve a drug? So let's talk about running a study versus marketing approval. Very different. So for running a study, if you're doing it in, say, Australia, you apply to the Australian regulatory authorities and the ethics committee for the study.

1:40:53And they do the review and request modifications and eventually approve. And then the studies run in Australia. If you want to then do a study in the US, you have to apply for an IND, just as you would if you were doing it any other time. But you include all the data that you got in Australia as well. And if you were doing it the other way around, it would be the same thing. And it's true for any two countries.

1:41:19Peter Attia:Meaning, if you had a drug that went all the way to the equivalent of a phase three ready for approval in Australia, and you come back to the U.S. from scratch and say, we want to be able to sell this drug in the United States, they're going to say, submit an IND. The FDA would. I think so, yes. Yeah, wow. But how much do you get the shortcut? Would they still make you do a phase one and a phase two, or would they let you go straight to phase three? I would think you could go right to a regulatory study. I mean, registration study. And in fact, this kind of thing is often done if, because typically you do the phase one study somewhere, but rarely in more than one, two or three countries.

1:41:57And then you can use that data to go to many countries for a phase two and then use that data to go globally. There are a few specific examples where you do have to run a phase one study before you go into that country. Best examples, the best defined examples are Japan. So to run a large study in Japan, you need to have run a phase one study in Japanese people. And there is a formal regulatory definition of who is Japanese from the Japanese regulators. and you have to provide that data before you can do a larger study in Japan. They're called ethnic sensitivity studies. And scientific rationale for this is that the genetic background of Japanese people can be a little different.

1:42:45Average body size is often different from people in the West and you want to make sure that the dosing and exposure will be safe and tolerable. But because the Japanese are so well organized and specific about what a Japanese person is, you can do these ethnic sensitivity studies in Hawaii, for example, or even California. Or you can do them in Japan. And I like to do them in Hawaii. China generally requires an ethnic sensitivity study also for the same reasons. And there are some specific examples of where there's toxicity of drugs in people on Chinese ethnicity. But they have a less specific definition of who's Chinese.

1:43:31Easiest way to do it is in China. All right.

1:43:35Peter Attia:So let's go back to BIMA. BIMA. You go into phase two now. By the way, at some point, doesn't Novartis sell this asset? Yes. So Novartis had strong confidence in Vimagromab. It was first in class, had really obvious biology in humans. And basically, Novartis ran maybe, I think, 16 phase two studies of one sort or another, or phase one, phase two study, different indications. Tried very hard. So the drug reliably and predictably increases muscle size, but not performance assessments in a major way. And I think that's because, remember in the rodents, in whom we saw both size increase and performance increase, the mass increase was large, 20 to 30 percent or more.

1:44:35In humans, it's 4 to 8 percent. And 8 is the absolute max.

1:44:40Peter Attia:Were those differences based on dose or starting mass? Biology. People are just not mice. Oh, sorry. I mean, the difference between the four and the eight, how much of that is dose dependent versus other demographic dependent on the pay? Like, you know, do you get more muscle mass in younger people, more muscle mass in people starting with more muscle mass? There's a trend to more in males versus females, a trend to more in younger versus older, but there's a lot of variability. Do we know if other variables such as resistance training, nutrition, protein consumption would have augmented these findings?

1:45:21Peter Attia:And how much were those variables controlled in these studies? So we know some of that. We try to control as much as we can. There was one study that has not been published in peer-reviewed form yet, but there is an abstract for it. If people want to find it, they can look it. So the BELIEVE study of Bimagrimab in obesity was just published a few months ago in Nature Medicine. If you look in there, this nutrition study is referenced. But there was a - We'll link to it in the show notes. Yeah. There was a study of Bimagramab in patients who were dosed at three different levels of protein, calorie, nutrition.

1:46:01And the bottom line is the more protein, and it was the recommended daily amount, half of that or one and a half times that I think. And basically within those boundaries, the more protein you ate, the more muscle you built. Probably shouldn't surprise anybody.

1:46:20Peter Attia:The other really interesting finding - And by the way, twice the RDA is only 1.2 grams per kilogram. Yeah, maybe that's what we used. It was 1.2. Yeah, I would argue had you gone to 1.6 or 2, you probably would have seen more hypertrophy. It's not been tested. I think you're probably right, but it hasn't been tested. It's interesting. So it suggests that in humans, you might've been substrate limited. could be amino acid limited or protein synthesis limited. It's a possibility. And not drug limited. I'll tell you a funny story about that in just a minute, but let's, but just to finish that study, the other very cool thing we found is that as you would expect, if you have half the recommended daily amount of protein calorie nutrients, you lost muscle mass, but the Magromab prevented that.

1:47:04Peter Attia:So there was, there was some, some biology working there for sure. No question. No question. So the The really interesting story is that when the Bimagrimab project, when it was still in the research stage, moved from Chris Liu's lab to David Glass's department, which I was part of as the clinical side of that, one of the things we really wanted to do was co-develop a nutritional component to this therapy for the exact reasons that you brought up. Yep. And at the time, Novartis had a nutrition arm. And so we were working with them to develop a specific nutritional supplement for what became Bimagramab.

1:47:43At the time, it didn't even have a code yet. But then Novartis sold their nutrition unit, I think, to Nestle. And so got the rug pulled out from under us on that side, and we were never able to fully pursue that. But in retrospect, I really wish we had.

1:47:58Peter Attia:Do you think this is a blind spot for big pharma? Just the role of nutrition and other behaviors that can potentiate drugs? Big pharma tries to control it, but they don't see that as their core mission. Yeah. But I'm saying a blind spot. I appreciate that they want to control it and that makes sense. But I'm saying like it's an opportunity lost, right? Probably. Like, here's a great example, right? Yeah. Like, Bima could have been more of a hit if maybe, and maybe not, but a drug like that could have been a hit had it been appreciated that, oh, by the way, like, you actually have to kind of do something to reap the benefits of this.

1:48:39We would have figured this out many years sooner if we had kept that nutrition element. But again, one of the challenges of big companies is there's so many people involved. They don't all know what the others are doing.

1:48:50Peter Attia:Yeah. Despite everyone's best efforts. So basically... It was a missed opportunity. And so did Novartis then spin it out after the less than expected findings in humans? So what happened was we saw muscle mass get larger, but not stronger. And parenthetically, that's the same as was seen with IGF-1 agonists and with androgen agonists. Remember, the SARMs were extensively studied, is that you can make muscles larger, they don't get stronger in the absence of resistance training. So it's not unique to the active and receptor antagonist pathway. And in a meta-analysis of the Novartis studies, where they looked at muscle hypertrophy and sarcopenia, the meta-analysis showed an increased six-minute walk distance.

1:49:42Nine meters. So a small effect.

1:49:46Peter Attia:That's my least favorite test in the world. It's surprisingly hard to standardize. Why wouldn't they just do something like a wall sit or something that really tests strength? Many other things were done. The timed up and go test, the short physical performance batteries. But six-minute walk was included in multiple studies, so you were able to do a meta-analysis of that. Got it. So an academic group did this, and they published it. And so the, the four to, I don't know, 8 % increase in muscle mass that these older adults got yielded a nine minute in nine meter increased in six minute walk distance.

1:50:30Peter Attia:I'm not convinced that's going to help anybody not fall. No, I don't think it will either. And, and, and Novartis didn't think so either. I guess I don't, I wasn't an insider at the time. I don't know why they out licensed it. I was the recipient of that. Yep. We were on the outside pulling. But the very last study Novartis did was a study in type 2 diabetics. Because we had had data that hemoglobin A1Cs decreased in patients given Bimagromab. And that study, which ran for 48 weeks, so it was 10 mg per kg monthly for 12 doses. Okay. So a lower dose than you were giving for hypertrophy? No, this maximized the - Oh, why did I think you said earlier 50 mg per kg?

1:51:19Peter Attia:In the phase one, we went up higher. You went up that high. Got it. We went up as high as we could because we wanted to know what would happen if people were overdosed later. Got it. Okay. And the answer was nothing. Okay. So at 10 mg per kg monthly over 48 weeks? Yeah, was a maximal dose in terms of effect size. and they saw the expected muscle mass increase. Interestingly, they saw a substantial fat mass decrease. And hemoglobin A1c in these type 2 diabetics decreased by about 0.7 % or 0.8%, absolute, which is a pretty good effect. Yeah, and do you think that that was on account of just more insulin sensitivity or was it a larger reservoir for glucose disposal?

1:52:04Both of those things. I think.

1:52:07Peter Attia:And these patients didn't have to do anything else. It wasn't like, in addition to that, they changed the way they ate or they exercised more. You gave them a drug that added muscle mass, took off fat mass, and lowered A1C by 0.7%. Yep. And they had standardized dietary advice to try the 500. So who, did Novartis run that study? Novartis did the whole study. And then they made a strategic decision that the effect size wasn't big enough. I mean, actually, I don't know what their strategic decision was, but they decided. The output was we're going to spin it out. Yep. The output was to spin it out.

1:52:42And at the time I was working with Joe Jimenez and Mark Fishman and Praveena Candula at Adidam Bio as an advisor. And we really wanted Bimagramab.

1:52:56Peter Attia:What year is this approximately? That it spun out was 2021. Okay. I think it was 2021. Yeah. Discussions had been ongoing in 2020, but I think it finally happened in 2021. Okay. So you guys acquired the asset, obviously, for a lot less than you could have produced it. Yeah. Still wasn't cheap because it was a phase two ready program. But we acquired - But you guys raised money for that acquisition? No, DEDEM did it all themselves. Okay. to their credit. And the plan was we were going to develop it in older adults with low muscle mass and impaired muscle function because we thought, who were also obese, because we thought this was the patient population most likely to benefit, losing fat and building muscle and maintaining muscle in the context of weight loss, we thought would be super important for those people.

1:53:58And remember, all of this happened in the context of nobody being interested in obesity. Everybody thought it was a wasteland for drug development. Every drug that had been developed in obesity had failed commercially. I mean, there are some that had been registered, right?

1:54:14Peter Attia:But by 21, you're saying pre-21. This was before NOVA's semaglutide data came out. Yeah, okay. That's right. Yep. And I'll tell you, I - Because that was 21, wasn't it? Later. and 21. Okay. So it's November of 21. Yeah. I think February 21, we started Versanus Bio, which is the company that licensed Bimagrimab from Novartis. So at that point, so I was the founding CEO. I was working with Elon Zipkin and we went out to raise money from investors because now, all right, great. We had this asset. We needed to run a big phase two study and we went out to raise money, I think we talked to 53 investors.

1:54:55Almost none were interested.

1:54:57Peter Attia:Because you told them indication, sarcopenia still. Well, it was sarcopenic obesity. Yep. And obesity was just not a successful area for drug development. So people weren't interested. How much did you need to raise? We ended up, how much did we need versus how much we got are different issues. Yeah, I know. That's why I asked. But we ended up raising$70 million. And what was your wish? If you could have had your wish list, what would you have raised? About$100. Okay. However, Atlas Venture and Medici liked the story. And I had worked with Atlas before. And Michael Gladstone was the partner. And it was Vani Marigi and Nick at Medici, Nick Williams.

1:55:46And we then built a investment syndicate and they funded the company. And then everything changed when Novo's data came out with semaglutide, which was amazing. It was the first really effective obesity medical therapeutic. But then when you're doing drug development, you skate to where the puck is going to be. It's Jay Bradner's favorite saying. But the puck was going someplace else now. Semaglutide was going to become the standard of Clare or some incretin agonist. We knew it. So what we did is we quickly repositioned the company to think about what is Bimagrimab going to do on top of that, because that's going to be the standard of care.

1:56:34So I quickly ran a bunch of mouse studies. and the efficacy was additive when you took Vimagrimab with semaglutide or terzepatide or liraglutide, it did them all. And the efficacy was sort of unprecedented, never seen anything.

1:56:53Peter Attia:For both fat loss and obviously for preservation of lean mass. Exactly. For weight loss, especially fat loss and preservation of lean mass. It was amazing. So the opportunity became much larger. And the board then, and I was part of the board, the board then brought in a super experienced CEO. This was Mark Brzezansky to lead the company then because we had a really big opportunity and we knew it. And we brought in a CMO, Ken Addy, because I had been serving as the CMO also. And then I stepped into president and CSO role just in terms of company organization. But we all kept working on the program.

1:57:35We ultimately ran what became the Believe Study. And the story here, we spent a lot of time thinking about what we would name our studies. The plan was Believe was going to be phase two, Become was going to be phase three, and Behold was going to be post-registration studies.

1:57:53Peter Attia:And then you actually have to come up with what those things stand for, knowing only what the B stands for when you start. I mean, this is so funny how drug name studies work. Yeah, you know how it goes. I know the drill. But they all became with B for B-magromap. Yeah. Now, you haven't asked me where B-magromap came from. So this is another interesting, wonky drug development thing. So the generic name is called the INN name for, I think, International Nomenclature. I'm not sure what it's an acronym for. Which is why it ends in Mab, obviously. So the suffix of a drug generic name is pre-specified based on the class.

1:58:30If you're the first in class, pick a new one. But the company gets to recommend the prefix and sometimes the infix. So Bima is the Indian god who's as strong as 10 ,000 elephants. And that's why bimagrimab is bimagrimab.

1:58:52Peter Attia:And then ends in MAB, monoclonal antibody. The grumab is a monoclonal antibody suffix. So any other drug that comes along in that class, what would be the nomenclature naming options? Antibodies are just complicated. There's a lot of different criteria for naming antibodies and you have options for infixes and suffixes. And what was the GLP-1 before liraglutide? Um, there's, there's a Xenotide. Xenotide. Right. So the tide became the thing that everybody had to link to going forward. Tide is peptide. There's so any, but they, they, were they forced into semiglutide, liraglutide, semiglutide, trzepatide?

1:59:33The tide is the, is used for that class, but there are other peptides that end in tide.

1:59:37Peter Attia:Yeah. Okay. So you guys ran believe. Yes. So BELIEVE, it was originally planned to be a 24-week study in patients with sarcopenic obesity. That was what we were going to do. But then when Novo's data came out and we got super excited about obesity, we said, oh my God, we got to do a bigger study. We're going to do it in combination with semaglutide. And this was during the pandemic. So there was a lot of complexity in supply chain disruptions. And remember, with semaglutide, it was a proprietary drug of Novo. We couldn't get the drug substance, so we had to use the commercial presentation of semaglutide.

2:00:22Peter Attia:Which was super expensive. Which was expensive, and it's an auto-injector. We couldn't make a placebo for that. So the study design included semaglutide as open label, but we did placebo-controlled Bimagromab because we had control of that. We used Bimagromab intravenously just because it was the fastest, most straightforward way to get into the clinic. Not public was that we were working hard on an auto-injector and we would have been ready in the next study for an auto-injector, but it was intravenous for Bimagromab. And then what combination should we use? We ended up doing something called a full factorial design.

2:01:05So we did all possible combinations of low-dose semaglutide, high-dose semaglutide, low-dose spumagromab, high-dose spumagromab, and placebo. So it's a nine-arm study. Why? Because we didn't know in humans what would happen with those different combinations. We didn't know if there would be adverse effects of the drug combinations or not. They did have a couple of adverse effects in common, diarrhea, for example. And this is in part why I did that pharmacology study in rodents, because if there was anything unexpected that would happen with the drug combination, I wanted to know about it. Technically, actually, and we had to do this because we weren't using the semaglutide in its intended population.

2:01:52I think it wasn't registered yet, I guess, was the issue. We had the data, but it wasn't registered, so it wasn't indicated in obesity. So if you're using an -

2:02:01Peter Attia:And so, yeah, you were using Ozempic, not Wagovi. You were still using - We used them both, actually. Oh, you did. Whatever we could get. Remember, they were in short supply. Yeah. And to the regulator's credit, they recognized all of this and were willing to allow us to substitute interchangeably Ozempic and Wagovi. The brands, yeah. So what were the findings of this six-month, nine-arm study? So it was originally going to be six months, but we changed it from the original plan with Bimagramab alone to the combination, and it eventually became 72 weeks of treatment, 48 weeks was the primary endpoint, and then we had a six-month follow-up period, so 104 weeks total study, two years.

2:02:47Peter Attia:Did you have to raise more money? We did. Yeah. I was going to say, that's a hard study to do for 70 million bucks. Yes, we did. And it was 500 people enrolled, roughly. 507 was the exact number. So. We found, so I guess number one, and one of the things I'm kind of proud of is I got the doses right. Because you wanted to see a partial response with the low dose, a full response with the high dose for both of the drugs, and see those kinds of dose responses in the combination arms. Remember, there's four combination arms, right? There's low-dose, BEMA, high-dose, BEMA, low-dose, SEMA, high-dose, SEMA, four combinations, and then placebo.

2:03:31And we saw the dose effects in all the arms. So that was good. And the primary endpoint was body weight. It wasn't what I wanted for a primary endpoint. I wanted waist circumference.

2:03:44Peter Attia:Why couldn't you get DEXA? Too expensive. We thought that since the registration decision is made on the basis of body weight loss, we wanted that as the primary endpoint. We included DEXA in every single patient. Yeah, it's crazy to me that you would be held to the standard of weight loss when in reality, a better outcome might be less weight loss. That's true. If you're preserving muscle. We were acutely aware of that. Yeah, that's awful. But it is not what the field was thinking at the time. No, I know, but it's just, I mean, it's good biology abuts regulatory simplicity for love. Yeah. Yeah.

2:04:27Peter Attia:That would be a charitable way. I really wanted waist circumference. And I wrote a long white paper about this because waist circumference is more closely linked to important clinical outcomes than is BMI or body. Yeah, for sure. Yeah. So body mass index is, if you're following longitudinally, is essentially the same as body weight. Because height doesn't change over the short term. Just for listeners. So we ended up using body weight as the primary endpoint. And it wasn't just regulatory intransigence. It was also what do investors and potential acquirers think. Everybody cares about what the approval endpoint is going to be.

2:05:09We wanted that to be the primary endpoint. But we measured all these other things. And to sort of zip ahead to the end, in the high-dose combination group, the body weight lost at 72 weeks was 22%, 23 % of starting body weight.

2:05:24Peter Attia:High, high. Yeah. Double positive. The high, high combination. And what was the highest? But the fat loss was 45.7 % of starting body fat. Now that's what you get with bariatric surgery. So this, to me, this is the first medical therapy that gives fat loss equivalent to or superior than bariatric surgery. That's amazing. Did you do any functional testing in that study? We did. We did, and we even did a preliminary observational study in overweight adults at different age cohorts, and we tested a few different things. We tested essentially timed up and go for short physical performance battery.

2:06:12We tested the 30-second chair stand test, which is my personal favorite, and we tested grip strength. but ultimately we went with grip strength in the BELIEVE study because it was the one most closely linked to clinical outcomes.

2:06:26Peter Attia:And did you see an improvement in strength? Small, but also it was a variable assessment. Okay. And it's in the published study that came out a few months ago. Now, Novo bought this asset from you guys, right? No, Lilly did. Oh, Lilly did. Yeah. Okay. So we were super excited about the study. It was ongoing. We were enthusiastic. We closed a series B in two tranches and we'd called the first tranche. And then the company got bought by Eli Lilly. And so they have Bimagrimab now. What are they doing with it? You have to ask Lilly. But they made some noise last fall that they were pausing the program or...

2:07:12They paused one study, but they still have other studies in clinicaltrials.gov, but you've got to ask them.

2:07:21Peter Attia:I see. So publicly, the only thing we know is they're still doing something with it, presumably testing it with terzepatide, I'm guessing, or retitrutide? Yeah, the study that's in clinicaltrials.gov is a complex combination study with terzepatide. I do think it's fair to mention the one adverse outcome that happened in the BELIEVE study that we weren't really expecting, which is an increase in LDL. Yeah, how much I remember that. Now, I thought that was actually something that Lilly saw, but that was in your study. Yeah. And how much of an increase was it? It was about 20%. Why do you think that was biologically?

2:08:00It's a direct effect of the drug in the liver.

2:08:03Peter Attia:So interesting. Again, you wouldn't expect this off-target, would you? This is on target. Because remember, there's active N-receptors everywhere. Oh, I didn't realize that. Yeah. Ah. So it's doing something to interfere with LDL clearance, presumably. I guess. But I don't know what the biology is. It hasn't been studied to my mind. Or no, maybe not. I mean, what would be a more plot? I mean, that would be studyable, right? Is it impeding LDL clearance or is it increasing LDL synthesis? I know who knows the answer to this. So Chris Liu, when he left Novartis, eventually founded a biotech called LAKNA in China.

2:08:41And he's made therapeutic antibodies to active and receptor type 2A, type 2B, and the combination, and he studied them. So he knows the answer to this, and I assume he'll publish it at some point.

2:08:53Peter Attia:And Lloyd, were there any adverse effects on glucose in the other direction? Did glucose ever go up? No. Okay. And did you continue in the believe to see glucose go down the way you did in the diabetic studies? Yes. Independent of what you would have seen from SEMA, I mean. Yes. So we have all of that data and it's published in the Nature Medicine paper. At EASD in September, which is a European meeting, we're going to publish the results of the six-month off drug results. So that's... That's off both drugs. Yes. And so the real issue is what's going to happen when you withdraw the drugs? And we know what happens when you withdraw semaglutide, right?

2:09:35Everything goes back towards where it was. It doesn't quite get there. And we're going to find out with bimagromab. I expect some things are going to reverse. Like we know that muscle mass with every muscle anabolic agent reverts towards baseline when you withdraw the therapy. I expect that's going to happen in the humans. It happens in the rodents. In Believe, we deliberately included patients with metabolic syndrome. So these are people who are pre-diabetic. So we can measure diabetic endpoints in these people. And it's going to be super interesting to see what happens there. Personally, if we had kept Vimagrimab in Bersanus and been a standalone entity, we would be well advanced into phase three by now.

2:10:20And the reason is because I believe, even with those LDL effects, which are not favorable, LDL predicts adverse cardiovascular endpoints. But I believe that—

2:10:32Peter Attia:You can monitor it and you can treat it. Yes. As an aside, since my personal interest is making drugs to prevent the most common causes of morbidity and mortality in older adults, side effect of that is healthy longevity. That's what I do. I am not making cardiovascular drugs, even though it is the number one cause of morbidity and mortality in adults in the U.S. and in many developed countries around the world. the reason is we've already got a lot of good drugs. We're just not using them for primary prevention, which we need to be doing more of. So with that aside, I would be well advanced in developing the Magrimab in phase three, but I think the paradigm for managing obesity is going to be induction and maintenance of remission, probably combination and injectable therapies to get people to their, you know, to move them categorically from obese to non-obese.

2:11:33And then they need something for maintenance, which might be something like orforglopron or some oral GLP-1 agonist to maintain appetite and satiety.

2:11:43Peter Attia:And you don't think just a lower dose of the injectable could do? Absolutely it could. Yeah. You're just saying economically, it might be easier to make it orally or something. Exactly. Yeah. Exactly. Okay. I want to pivot and talk about one other thing, which is also an area where you know a lot about it, which is selective mTOR inhibition. We're not going to spend as much time on it, of course, but again, talk to me about where your head is at these days on that pathway in general. Do you believe that this is gyroprotective in humans? I mean, it's been well-established how gyroprotective this is in mice.

2:12:21Peter Attia:Almost assuredly, I think it'll end up being geoprotective in dogs. So it might be safe to say that inhibiting mTOR in everything from yeast to dogs and maybe even primates extends life. We don't have a clue if it's going to in humans. We'll never probably get to directly test it. There are really good, compelling arguments on both sides of why it may or may not be the case in humans, including the longevity quotient argument and things like that. What are your thoughts? I think it probably will. It's highly conserved biology across evolution. So I think so. Reductively, if you envision mTORC1 as a master regulator of sensing, sort of integrating nutritional inputs and then deciding to grow or not grow, and not grow means circling the wagon upregulating autophagy and recycling pathways?

2:13:19I think it probably would. I think the effect size is going to be modest. Is it just as it has been preclinically? And I think it's TORC1. I haven't seen a lot of new data on that.

2:13:32Peter Attia:Well, there was that study somewhat recently suggesting that rapamycin impaired, I don't know if it was impairing MPS or some other metric of physical performance or something. Obviously, you're familiar with the agents you've tested. You know, we're still at sort of the infancy of these drugs, right? What do you think is standing in the way of more drug development on more and more selective, potentially higher efficacy, but potentially lower side effect burden versions of drugs that can inhibit mTOR complex one? the selectivity is the big challenge because with rapalogs, as you know, there's their TORC1 selective, but there's a down regulation of TORC2 with sustained exposure.

2:14:23And I don't know that we, you know, so in RestoreBio, we tried to manage that by a combination of a catalytic and an allosteric inhibitor, which seemed to do it. And I know there are other companies that are working on other ways to get TORQ1 selective inhibition. And I think that's what we need is a real TORQ1 selective inhibitor. And then we can test the biology.

2:14:45Peter Attia:And do you think that just intermittent dosing of Everolimus or Sirolimus gets that? Maybe. Again, it's hard to tell in healthy people because, you know, in cancer, when you study mTOR inhibitors, the cancers have a highly upregulated pathway and it's easy to see the biology. You can't really see the active biology in humans measuring blood. It is not necessarily the tissue you want anyway. When we do this in rodents, we measure their liver activity. And I think we mentioned, you and I discussed this before, that in young rodents with fasting, they down-regulate mTOR, as you would expect. In old rodents, they didn't.

2:15:26So it makes me call into question the whole concept of intermittent fasting in older people, because I don't know if it'll do the same thing. Yeah.

2:15:37Peter Attia:Again, imminently testable. Nobody's lining up for liver biopsies though. No. And we can't get that with MRS or anything else. I don't think so. It's just... Yeah. Again, the price of admission is so great on that. Okay. Final question slash topic. Like, as you think about drug discovery over the next decade, I'm not going to ask you the question everybody's thinking is, how is AI going to help? We'll punt that for now. Thank you. Yeah. What are you most optimistic about in terms of pathway, disease? Where are you most excited? Where do you think we're going to be in 10 years where there's been a step function change?

2:16:20I think we're starting to wake up to the concept of real medicine, real medical prevention. I mean, this is something I've been saying for years. You've talked about it a lot, is that we need to get away from being a sick care system to a health care system. And the way you do that is with preventive medicine. And the way to implement it is you need better primary care and you need codes for preventive visits because right now they're, if I wanted to see a patient for prevention of cancer, for example, so my new company is Cancer Prevention. There's not codes for that. So you can't bill for it.

2:17:00So there's, there's a lot of institutional hurdles that we need to get through. But I think people are waking up to the concept of, I want to stay healthy rather than get sick and get treated.

2:17:14Peter Attia:Say a bit more about your current company and how could one develop a drug for cancer prevention? Yeah. So this is conceptually difficult to wrap your head around. But again, where do new drugs come from? They come from reading the literature and thinking, which is sort of what I did after Versanus ended for me. It's still ongoing in Lilly. And there are some papers published over the past 5 to 10 years about drugs that cause cancer. So if a drug causes cancer, it must be most likely inhibiting a cancer-protective pathway. Most drugs are inhibitors of things. The prototype for this is serafinib, which is a multi-kinase inhibitor that's used to treat renal cell carcinoma and hepatocellular carcinoma primarily.

2:18:06If you give that drug to people, about 10 % of the older patients get skin cancers. Why is that?

2:18:13Peter Attia:And are these melanomas or are these squamous or basal cells? The cancers they get seem to be the prevalence that's reflected in the normal population. So almost everything that's ascertained is basal cell and squamous cell. And more recently, we understand the pathway biology of that. Serafinib is a multi-kinase inhibitor. It inhibits a lot of kinases. But one of the ones that it inhibits is the sensing kinase that triggers something called ribotoxic stress. And this is a pathway that causes cell death. That pathway, if you turn it on irreversibly and covalently, is the target of some of the nastiest toxins that you know about, like diphtheria toxin, sarsin, ricin.

2:18:59It's a very, very potent pathway. My innovation is putting together different parts of the literature. I came up with a way to turn it on in a gentle and controlled fashion. Remember, it's on constitutively in people, because if you turn it off with these multi-kinase inhibitors, you get cancer. So the hypothesis of the company is if we turn that pathway on a little more, we'll prevent cancers. Not all of them, but maybe 50 % is what I'm hoping. And since skin cancer is almost as common as all other cancers put together, we've got to start there. But in a phase two study of older adults, and older adults in this context means 50 and up.

2:19:46Sorry, Peter.

2:19:47Peter Attia:I'm in that category squarely. Don't worry. I have been for a while. And who have had at least five skin cancers in the past. Those people have a 50 % chance of having another skin cancer within a year. So if we recruit a cohort of 100 or 120 of those, we would be able to test the low dose, high dose, placebo, and actually measure cancer prevention in a phase two study. Now, is there a risk that it will only work in preventing squamous cell and basal cell carcinoma, but will not progress an epithelial tumor or prevent, I'm sorry. That's possible because the only data we have are for skin cancer.

2:20:26But even if it only prevented skin cancer, that's a really big medical need. But I think it will work on multiple cancers. But it's going to be almost impossible to test that before approval just because cancer incidence is a really rare event.

2:20:42Peter Attia:Yeah. Yeah. I guess the next thing that would be an interesting question, Lloyd, would be you take a bunch of patients who have successfully undergone adjuvant therapy for a stage three epithelial cancer. So I would think colon cancer or breast cancer. They're NED, no evidence of disease for the listener, but there's a 50 % chance they're going to have a recurrence. You know, you stratify it in a way that you basically find people who have a very high risk of a cancer recurrence, and then you treat them. I think that's one way to do the other study. What I've done in COSLEP Therapeutics is a collaboration with the Broad Institute, where we took our tool compound.

2:21:31Now, we don't quite have a development candidate yet, but we took the tool compound, which is good enough, and run it through their panel of a thousand cancer cell lines to see what tumors are sensitive to it. So this would be a treatment mode rather than a prevention mode.

2:21:48Peter Attia:And you think it could have efficacy in treatment as well? That's the question we were asking. Okay. And melanomas emerged as by far the most sensitive tumor. Now, I'm not sure why because my hypothesis, the thing about skin cancer is it's got a heavy mutational burden because of all the UV exposure. It's the highest mutational burden organ we have in normal people. And I thought that was going to be it. And there was a correlation between mutational burden of the cell lines and susceptibility to this mechanism, but it wasn't great enough to explain the tumor susceptibility. So it's something else.

2:22:27That would be good news. Yeah, it's there, but it's not good enough. So I don't know why melanomas are so sensitive, but they're enormously sensitive. So I'm very confident this will be a therapeutic for melanoma as well. And it augurs well to the idea of preventing melanoma, which is also testable and has been proven with a therapeutic intervention in a wonderful study conducted in Australia. The intervention was intensive sunscreen use compared to usual practice. So it's going to be testable in a large phase three study, but not before that. So, yeah. So that's what I'm doing. And I think preventing cancer -

2:23:09Peter Attia:That's a very interesting idea. I mean, talk about a whole new playing field, right? Yep. Pharmacologically. Yep. Nobody's made a drug for this mechanism. Because we usually think avoiding cancer or preventing cancer comes down to avoiding carcinogens. Which we should definitely do. Yes, yes, absolutely. So don't drink alcohol much. Or don't smoke. Don't smoke. Be as insulin sensitive as possible. Yes. Yeah, all of these things. All of those things. Lose weight if you're overweight. We know that successfully treating obesity prevents a bunch of cancers. Yep. And we know that from the Swedish obesity study, which is an observational cohort of Swedish patients who've undergone bariatric surgery.

2:23:54And these people are being followed for decades. needs, it's doing all the good things you'd expect of successfully managing obesity.

2:24:04Peter Attia:Well, Lloyd, this has been great. This has been kind of a wonderful education on drug discovery using, I think, a very interesting drug in BEMA as a case study for the complexity and the nuance of the process. And by the way, I don't think I realized that the BEMA story is still ongoing. So that's great. So we're going to continue to follow this biology and it'll be interesting to see where Eli Lilly goes with this drug. But it sounds like based on what's showing up on clinicaltrials.gov, they're following in your footsteps in that they're probably testing this in parallel with the newer generation GLP-1 agonists.

2:24:40As best I can tell, that's what they're doing. And it's not now just Eli Lilly because many other companies who have seen the believe data, because we've been presenting it at national meetings and international meetings, for that matter. There's a lot of other pathway inhibitors that are under development. Yeah.

2:25:02Peter Attia:Well, really appreciate your time, Lloyd. It's been a pleasure to chat again. Great efforts. Yeah. Thank you. Thank you. Thank you for listening to this week's episode of The Drive. Head over to peteratiamd.com forward slash show notes if you want to dig deeper into this episode. You can also find me on YouTube, Instagram, and Twitter, all with the handle peteratiamd. You can also leave us review on Apple podcasts or whatever podcast player you use. This podcast is for general informational purposes only and does not constitute the practice of medicine, nursing, or other professional healthcare services, including the giving of medical advice.

2:25:40Peter Attia:No doctor patient relationship is formed. The use of this information and the materials linked to this podcast is at the user's own risk. The content on this podcast is not intended to be a substitute for professional medical advice, diagnosis, or treatment. Users should not disregard or delay in obtaining medical advice from any medical condition they have, and they should seek the assistance of their healthcare professionals for any such conditions. Finally, I take all conflicts of interest very seriously. For all of my disclosures and the companies I invest in or advise, please visit peteratiamd.com forward slash about where I keep an up-to-date and active list of all disclosures.

2:26:23And I'll see you next time.