255-Latest therapeutics in CVD, APOE’s role in Alzheimer’s disease, familial hypercholesterolemia

1. 0:00 – 0:11

   Intro

   1. PAPeter Attia0:01

      [upbeat music]

2. 0:11 – 6:07

   Why familial hypercholesterolemia (FH) matters: prevalence, stakes, and episode roadmap

   1. PAPeter Attia0:11

      Hey, everyone. Welcome to the Drive Podcast. I'm your host, Peter Attia. This podcast, my website, and my weekly newsletter all focus on the goal of translating the science of longevity into something accessible for everyone. Our goal is to provide the best content in health and wellness, full stop, and we've assembled a great team of analysts to make this happen. If you enjoy this podcast, we've created a membership program that brings you far more in-depth content if you want to take your knowledge of this space to the next level. At the end of this episode, I'll explain what those benefits are, or if you want to learn more now, head over to peterattiamd.com/subscribe. Now, without further delay, here's today's episode. [upbeat music] My guest this week is John Kastelein. John is a genetic researcher and clinician scientist known for his work in the field of familial hypercholesterolemia and the development of lipid modulating drugs. He is currently a professor of genetic medicine at the University of Amsterdam, where he leads the Department of Vascular Medicine. John has been the main driving force behind the development of a treatment for homozygous familial hypercholesterolemia, a severe form of FH. Now, some of you may be listening to this saying, "What the heck are you talking about?" Well, it's important to understand that FH or familial hypercholesterolemia is the second most common form of hereditary heart disease, right after elevated LP little A. And you know from probably listening to previous versions of this podcast that elevated LP little A is staggeringly prevalent in the population, and by extension, therefore, so too is various forms of FH. And while we use FH as an important place to start this discussion because it becomes an important way to understand therapeutic options, the subject matter that we cover here is, of course, applicable to anybody who's interested in minimizing their risk of cardiovascular disease. John has also led several clinical trials, including the pivotal ODYSSEY Long-Term and ODYSSEY Outcomes studies, which helped to establish the safety and efficacy of PCSK9 inhibitors in the treatment of FH and other forms of hypercholesterolemia. As I said, we start the discussion by focusing on familial hypercholesterolemia. We talk about what it is, how you define it, how you can be aware if you have this, what the genetics are that underpin it, what we do with kids that have this, et cetera. We then talk about the history of CETP inhibitors, which is indeed a sordid history. These have been a class of completely unsuccessful drugs that have resulted in much hype and fanfare without any tangible results. However, John makes a pretty compelling case for the most recent version of these drugs to be not only a potential game changer, a word that I hate, for cardiovascular disease, but perhaps even more interestingly, for Alzheimer's disease and type 2 diabetes. This dovetails very nicely into our final topic of discussion, which is that in the role of APOE. Now traditionally, when you hear me talk about APOE, I'm talking about the gene APOE and its three isoforms, APOE2, APOE3, APOE4. It's important to understand that, of course, those genes code for a protein that goes by the same name, apoE, although it is not fully capitalized, and that's how when you're reading it you know the difference. What we talk about in this episode is what that actual protein does, and why is it that someone with the APOE4 gene codes for a version of that protein, which by the way only differs in one amino acid from the one coded for by APOE3. And we talk about why the protein that is coded by the APOE4 isoform produces a much greater increase in the risk of Alzheimer's disease and cardiovascular disease. What's most interesting to me about all of this is that it ties back very nicely to the discussion of how this most recent CETP inhibitor might work. And what I'm left with is a sense of profound optimism that sometime in the next five years, we may indeed have therapeutic molecules that we can use specifically for high-risk patients such as those with APOE4. So overall, I would say this discussion surprised and delighted me much more than I expected, and I know that even though it's a technical topic, it is something that is going to be of great interest to anybody who cares about heart health and brain health. So without further delay, please enjoy my conversation with John Kastelein. [upbeat music] John, thank you so much for staying up late into your evening in Amsterdam to make time to speak with me. This is a podcast that really came across my radar courtesy of one of my mentors, Tom Dayspring, who basically got me interested in the work you were doing and, uh, said, "Look, if you'd like to speak with John, uh, we might be able to twist his arm that, to make time for this." And so I'm both gracious that Tom got me into your work and, and that he was able to convince you to, uh, to sit down with us.

   2. SPSpeaker5:07

      So am I.

   3. PAPeter Attia5:08

      As you know, because you, you mentioned to me earlier that you've listened to the podcast, you're probably familiar with how much we talk about cardiovascular disease. It's not really an accident, right? I mean, it is the leading cause of death globally. I've-- I don't think you can state the stats enough, right? I mean, the last time I looked, 19 million people died in the world due to ASCVD, and the second place killer was cancer at something like 12 to 13 million. So it's not even close that what we're talking about is the leading cause of death, and it's for that reason I think that we need to make sure that we take every opportunity to educate people about this. And I want to start with a genetic condition called familial hypercholesterolemia, which is probably far more prevalent than people realize. In other words, there are thousands of people listening to this podcast who are affected by that. So tell us what FH, as it is abbreviated, is.

3. 6:07 – 7:25

   Kastelein’s origin story: building large-scale FH detection in the Netherlands

   1. JKJohn Kastelein6:07

      Yeah, so we were the first in Holland who actually had a very large-scale organization to find these people. So I trained in Vancouver. I was a visiting professor also in Vancouver by Michael Hayden. I had my Canadian exam, so I did lipid clinic in Vancouver in a time when there were no lipid clinics in Europe or hardly. So when I got back in Europe and I started the lipid clinic in Amsterdam, about 60% of all my referrals were FH, and I thought, "Why would that be?" And then when you go back into the history of my country, my country is about the size of Rhode Island, and it has 18 million people. These people have not really moved in the past, so there are large provinces with consanguinity. It's not consanguinity as, for example, with the French Canadians or the South Africans, but it's nevertheless consanguinity. So we got a huge €30 million grant from the Dutch government to actually start actively find these people because when I started doing this in our large lipid clinic in Amsterdam, we still had mortality or severe anterior MIs at ages between 20 and 30 in men, for example.

   2. PAPeter Attia7:24

      Wow.

4. 7:25 – 9:38

   What FH looks like clinically: inheritance, early onset, and physical signs

   1. JKJohn Kastelein7:25

      So I can almost remember all of them, people that were fit like you but didn't know they had an LDL of 300 mgs per deciliter from their birth onwards and actually were playing some tennis, and then they got a massive anterior MI. And in those days, there was not much stenting yet, and so the only thing you could do is put someone in a, in a cardiac care unit and actually do all kind of vague, superficial things. So FH is a true autosomal dominant disease, meaning it, it's not sex-linked. Uh, you don't need two parents to get it. You only need one parent to get it. It's almost 100% penetrant, meaning that if you have a robust mutation in one of the genes that cause FH, you're almost certain to get the phenotype, and the phenotype starts very early. We have in Amsterdam the largest pediatric lipid clinic in the world. We have seen over 2,000 children with heterozygous FH, so we know that this disease starts very early in life, then becomes symptomatic in your teenage years in the sense that the cholesterol that is elevated in the circulation starts depositing on your tendons, especially the extender tendons on the dorsum of your hand and in your Achilles tendon. And then also you start very slowly seeing the arcus, so the arcus in the eye and sometimes the deposits, uh, on your eyelids, the xanthelasmata. And so those are the physical manifestations of the disease. And then very often, especially in the old days, the first really serious manifestation is either angina or a heart attack. And one of the things that is so dangerous about this disorder is that it's the plaque that you get in FH is a soft plaque. It's a cholesterol-rich large plaque that is very often proximal in the coronary arteries, meaning that if a plaque like that bursts, you either have an, an occlusion of your entire LAD or you have a main stem occlusion, which kills you right on the spot. Uh, you know, sudden death, actually.

5. 9:38 – 15:38

   Phenotype vs genotype: how FH is diagnosed and why adults are harder than children

   1. PAPeter Attia9:38

      Those sort of proximal left main or distal left main are we-- it's referred to as the widow maker. I, I want to just back up and make sure some of the stuff that you said makes sense to people. So you said so much there that has my eyes wide open, and I understand this stuff, so I just want to make sure everybody gets it. So a couple things. Let's differentiate when you use the word phenotype and genotype, let's explain to people how one defines the phenotype of FH. In other words, what is the objective metric by which we define this condition?

   2. JKJohn Kastelein10:13

      So the objective metric, as with all patients, is history first, and very often in these families, there is a family history of premature coronary disease. That's one of the first things that speaks for these families. Then second, these people have elevated LDL cholesterol without any other abnormality. So HDL is normal, triglycerides is normal, LDL is elevated, and you need to find elevated LDL in first-degree relatives also, so in children of that individual or in one parent and in siblings. And with that LDL, you can reasonably construct a family tree where you know who's affected and who's not. But unfortunately, as you undoubtedly know, there's overlap between unaffected family members and affected family members in LDL cholesterol. That's why we decided a long time ago to go for the mutation, so to go for the genotype because for- of course, for the genotype, there is no overlap. Either you carry the mutation or you don't carry the mutation.

   3. PAPeter Attia11:21

      Now, this is where we have a bit of an issue, right, John? Because unlike, for example, Lp (a) , which we've had several podcasts on, where the phenotype is unambiguous, you have an elevated level of Lp (a) , the lipoprotein, the genotype is also very clear. It's a one-to-one mapping, right? LPA is the gene that codes for apolipoprotein little a, and away we go. Here we have a very heterogeneous genetic set of causes. In fact, I could be wrong on this, so please correct me. I believe I've read that there may be over 3,500 different mutations that would roll up into FH, i.e., that would produce this phenotype of... And are we using, by the way, an LDL-C cutoff of 190 milligrams per deciliter?

   4. JKJohn Kastelein12:11

      Yes, that's what we are mostly using, 190. Of course, if you start a lipid clinic and you get referrals, there's referral bias, so the initial patients that I saw all had LDLs of 300 or more. Now, the genotype is very interesting. So what we've done is, of course, if you start determining a genotype, you have to be absolutely sure that your diagnosis is correct because if your diagnosis is not correct and you don't find a mutation, it's meaningless. So what we've done is we've used the children as our kind of diagnostic linchpin because if a child has high cholesterol, with the exception of primary hypothyroidism, there's almost no other cause for elevated LDL cholesterol in a child than FH.

   5. PAPeter Attia12:59

      Right. And it's important to explain, I think, to people that although you've alluded to it, I want to make sure people understand, there are a lot of people walking around with an LDL cholesterol of 200 milligrams per deciliter.

   6. JKJohn Kastelein13:09

      Oh, yeah.

   7. PAPeter Attia13:10

      But they might have triglycerides of 300 milligrams per deciliter. They might have type 2 diabetes. They might have thyroid disease that is untreated. They might have renal disease that is untreated. There are lots of other diseases for which a side effect is elevated LDL cholesterol. And what you're saying is, "Look, we've got to rule those out," right?

   8. JKJohn Kastelein13:32

      Yes. And therefore, we started in the children, Peter.

   9. PAPeter Attia13:35

      That's my point, yes, because in the kids, there's far less likely to see those other diseases.

   10. JKJohn Kastelein13:40

       Yeah, it's almost in all the thousands of kids we've seen were just a handful had, had a secondary cause for elevated LDL. The vast majority actually all had a genetic cause, and we've published this in JAMA and in Lancet. So when we, we had a cohort of 220 children where we had three-generation family of elevated LDL, so there was no doubt that they had FH, then we started sequencing. We could find a mutation, I mean, next-generation sequencing, everything, exon-intron, the whole, in 95% of cases. So that is a very different number than what you normally read in the literature for adults, where people can find 50, 60% at the max, and they have no explanation for the other 40%, telling you that the diagnosis, the clinical diagnosis is not that good yet in adults, unfortunately. But it's very good in children, 95% mutation. So out of those 220 kids, 5%, which is 11 kids, we also have never, ever found a mutation, and we've really tried. So that tells me that there are more genes-

   11. PAPeter Attia14:59

       There are mutations we don't know about yet.

   12. JKJohn Kastelein15:01

       Exactly. Now, the 95%, if you make that 100%, then it's 95% the LDL receptor, 4.5% ApoB, and 0.5% PCSK9 gain-of-function mutations. So that, in a country like ours, is the division between the different genes. LDL receptor, vast majority-

   13. PAPeter Attia15:25

       And here it's, it's not that profound, right? I mean, here in the US, I was always under the impression it's probably, call it 70, 80% is LDL receptor.

   14. JKJohn Kastelein15:35

       It's probably more like that in the US, yes.

6. 15:38 – 19:25

   Mechanisms behind FH genes: LDL receptor, ApoB binding, and PCSK9 gain-of-function

   1. PAPeter Attia15:38

      And let's explain to people what's going on here. Again, I think it's worth understanding the pathophysiology because we'll get into treatment. But when we say mutation in LDL receptor or mutation in ApoB or gain-of-function in PCSK9 protein, explain briefly what each of those means and why would each of those translate to the higher biomarker that people are used to seeing, right? Everybody listening to this knows what their LDL cholesterol is, but why would these mutations lead to 3x normal levels of LDL cholesterol?

   2. JKJohn Kastelein16:11

      That's actually a wonderful and very easy-to-understand story. So every LDL particle has an ApoB protein kind of cringled around it, almost like a snake, and one area of that protein is sticking out of the LDL sphere, and that is a binding domain. That is around amino acid 3,500. We actually know that. Then there is a receptor for that particle sticking out from your liver cell, your hepatocyte, and these two bind normally. So the LDL receptor grabs the LDL particle, and then the whole complex is internalized into the lysosome, endosome, where it's basically dealt with. Now, next to the LDL receptor on the hepatocyte surface sits another protein called PCSK9, and that protein degrades the LDL receptor, and that is the balance in nature because you can't have an, at least not in the old days in evolution, you can't have an overactive LDL receptor because then you clear every LDL particle from your circulation. So you have to have a balance, and the PCSK9 protein that degrades the LDL receptor gives that balance. But if there is a mutant in the binding domain of ApoB, you have something that we call FH or familial defective ApoB, which is basically the same. If you have a mutation in the LDL receptor, you can't bind the ApoB. And then if you have a very active PCSK9 and you basically degrade all your LDL receptors, the end result is the same. There's not enough LDL receptors for the LDL particles. So all three things converge at the surface of your liver cell, and the problem with either one of the three always leads to elevated LDL cholesterol, and that leads to all the downstream things that we just discussed.

   3. PAPeter Attia18:10

      So for folks who like to anthropomorphize things- You can picture the LDL with the ApoB wrapped around it as a baseball. The LDL receptor is a baseball mitt sticking out from the liver, and the PCSK9 protein is something that smacks the mitt and closes it. And so basically, one form of mutation is mutations that change the shape of the baseball mitt so it can't catch the ball or it catches it very poorly. Another mutation changes the shape of the ball so the ball doesn't fit in an otherwise perfect baseball mitt. And then the final mutation is one that makes too many of the things that swat the baseball mitts.

   4. JKJohn Kastelein18:52

      Absolutely. But if you make the diagnosis right in a child with elevated LDL cholesterol, you've excluded the rest. You know his father has high cholesterol, his grandpa had a high cholesterol, one sibling has a high cholesterol, you know, yeah, well, this is FH. Then you do a mutation screen, you find in 95% of cases you can find a mutation. The vast majority is the LDL receptor, then comes ApoB, and then comes PCSK9. So in our country, PCSK9 mutations gain-of-function are very rare, .5%.

7. 19:25 – 22:10

   Beyond classic FH: plant sterols, ABCG5/8, and why advanced clinics see overlaps

   1. PAPeter Attia19:25

      You know, another mutation that we've seen even in our practice, and we have a very small practice, but I suspect we also disproportionately collect people who are higher risk. But I've seen two cases of what appears to be ATP binding cassette G5G8 loss of function. So we see people who have FH, and I mean, I can't rule out LDL receptor defects or gain-of-function PCSK9 because we're not gonna do that genetic test, but what we can see is that they have levels of sitosterol and campesterol that are more than a log fold higher than the 95th percentile. And from that, I don't know, would you agree that we're imputing it does change our management because ezetimibe becomes first line, but more importantly, just as a very curious finding, would you agree with our likely inference that that's probably the driving mutation in those people?

   2. JKJohn Kastelein20:24

      I would completely agree, and I'll take one little step back. So mutations in these three genes that we just described are very prevalent in the general population. In fact, we have calculated that they are probably like one in 250, which makes FH by far the most frequent autosomal dominant disorder in men. Now, the Koreans and a number of other people have actually looked in children with high cholesterol at sitosterol and campesterol levels. As you know, they should be like 12 for sitosterol and 14 for campesterol. However, if you put the level at, like, 35, they find a preponderance of kids who have increased plant sterols in their circulation, and they find mutations in ABCG5G8. So in the old days, we thought that sitosterolemia, which is the disorder associated with ABCG5G8, was like one in a million. That is probably a huge underestimation, and based on a number of studies in different countries, we now think it might actually be like one in 150,000 or so, 10 times more frequent than we originally thought. And what we also didn't appreciate is that if you are heterozygous for a loss-of-function mutation in ABG5G8, you also have, exactly as you said, increased sitosterol and campesterol levels. And so very likely these disorders, especially in an advanced clinic like yours, will coincide quite frequently because they are not rare. They are actually both not that rare. So I have exactly the same experience.

8. 22:10 – 30:01

   Why cholesterol deposits show up in tendons and eyes: xanthomas, arcus, and clinical pattern-recognition

   1. PAPeter Attia22:10

      Let's go back a little bit to this clinical presentation. We talked about it, but I, I think some of these physical signs might be surprising to people that go beyond the obvious, which is atherosclerosis. So you mentioned tendon xanthomas. You mentioned cholesterol deposits elsewhere in the body. Explain a little bit more about if we understand why does cholesterol tend to accumulate in those particular areas, for example, extensor tendons over flexor tendons. Do, do we have a sense of why that's the case?

   2. JKJohn Kastelein22:40

      The theory is that it's linked to movement, Peter, so that tendons that are used very, very frequently, that there is around those tendons, there is a preponderance of macrophages and monocytes. These monocytes, macrophages are capable of storing LDL cholesterol, and when there are enough of them, you actually see it. You physically see it, and you can feel it also. So the two most frequent places where you'll find these deposits are the extensor tendon of your hands, which you use the whole day, and of course, your Achilles tendon. But in my career, which as you can see on my gray hair has, has been a while, I've seen xanthomata also in the patella, and so that tendon actually on the tibia. I've seen them under wedding rings, so actually a xanthoma under a wedding ring. I've seen them everywhere. And then the deposits on the eyes are probably also linked to movement because you're blinking your eye the entire day. And then you have a deposit of cholesterol in your cornea, which is called an arcus cornealis. So it's not in your lens, it's in your cornea. Why that is, there is a lot of conjecture about it, but I don't think people really know. But sometimes, actually, I have made the diagnosis of FH once in a KLM airplane. So I was kind of lying. There was a stewardess who wanted to give me a drink, and she had very blue eyes, you know, these Dutch blue eyes, and I looked in her eye and I saw a ring I said, "When you go back, you need to have your cholesterol checked." And she proved to have FH. So sometimes you can make the diagnosis by really paying attention in the subway or et cetera, because also the extended tendons on your hands are sometimes really visible. There are a number of old Dutch paintings from the 17th century where you can still see the tendon xanthomas on the hands.

   3. PAPeter Attia24:46

      Wow. And of course, they obviously had no idea what was causing that shape, but it was, they wanted to be as accurate as possible and they painted it. So we've established how the diagnosis is made. Do you require the clinical manifestations for the diagnosis? In other words, if you have a 50-year-old person who has had elevated LDL-C for as long as they've had blood tests, so they would say to you, "Yep, you know, Doctor, going back into my teens or 20s, I had LDL-C north of 200 milligrams per deciliter. My doctors always said that's too high, and I didn't care," blah, blah, blah. But now he's standing in front of you and you examine him and he has no evidence of xanthomas, no arcus, nothing at all, no physical sign whatsoever. You send him for a CT angiogram and his coronary arteries are clean, but lo and behold, his father also has elevated cholesterol. Let's say he died of heart disease at the age of 78, so he died, but not prematurely. Absent a genetic test, which I assume would seal the deal, do you say at least phenotypically he still meets criteria for FH? And also, let's assume that you've ruled out every other medical thing, so he doesn't have hypothyroidism or kidney disease or insulin resistance.

   4. JKJohn Kastelein26:06

      You don't need the physical stigmata to make the diagnosis because we've seen that time and time again. There are people, for some reason, that have very elevated LDL cholesterol. It is genetic, it is dominant, it is, but it doesn't lead to the physical stigmata. What I haven't seen that many times is what you're describing, is that the CT angio is normal in a 50-year-old. So if someone tells me that they had in their teens or late teens or when they went into the military, you know, when they're 18, they had an elevated LDL cholesterol, and when you then see them when they're 50 and they haven't been treated to have a normal calcium scoring and/or a normal CT angio, that is really rare for LDLs, let's say, you know, between two and 300. That's really rare. But we've seen them, and it's interesting is that, so there is a cholesterol overlap. It's like a Gaussian distribution for people like you and me. Then there's also a Gaussian distribution for heterozygous FH. There is an overlap between the two, but now we know, and Evan Stein and I have actually published about that, is that there's also an overlap between homozygous FH and the end of the distribution of heterozygous FH. So you can never use LDL alone as 100% certain marker for your normal, you have heterozygous FH or you have homozygous FH. This is a syndrome diagnosis, so you need family members, history, and arcus and this and that. And then what's also very important is what you said in the beginning, that this is a unique LDL disease. The moment there's elevated triglycerides, for example, and low HDL, then you immediately have to think about something entirely different, especially in a individual that's not obese or diabetic.

   5. PAPeter Attia27:58

      Let's go back and make sure we explain to people the difference between what I assume is much more common, which is heterogeneous FH and homogeneous FH. Can you explain to folks what that is?

   6. JKJohn Kastelein28:11

      The terminology I find sometimes difficult. So we know, and I'm sure you've discussed that also, I think with Dan Rader, is that there is called polygenic hypercholesterolemia. That sometimes is so severe, the polygenic hypercholesterolemia, that it looks a lot like heterozygous FH, and some people even say if you've had those genes also from birth onwards and you've had high LDL cholesterol, you know, from a young age, the risk of that severe polygenic hypercholesterolemia is just the same as for heterozygous FH. I still find that difficult because I see when you go look at premature MI, for example, patients with premature MI, there's a huge enrichment for real heterozygous FH, so the homogeneous heterozygous FH. The heterogeneous FH or polygenic hypercholesterolemia, whatever you wanna call it, is in my view still a less severe clinical picture. I've never seen people with that side of the spectrum get heart attacks in their 20s or 30s. So we have found 25,000 heterozygous FH patients, yeah, we've published on this, so it's a humongous database. And the monogenic, especially LDL receptor gene mutations, and especially if they're severe, so a premature stop codon, so you don't have any protein at all, that is the most severe form of inherited hypercholesterolemia in my dictionary and in my experience. But truth is, is that your polygenic or heterogeneous FH is probably much more common, even more common than heterozygous, the monogenic form of heterozygous FH.

9. 30:01 – 34:16

   FH diagnostic criteria and pediatric treatment thresholds: Dutch Lipid Clinic scoring and early statins

   1. PAPeter Attia30:01

      I want to come back to this distinction when we start to talk about therapy in as much as it changes either the initial steps we take or the expected number of steps we take therapeutically. But before we do, I want to put a, a bow on a few other things. Can you formally state again the Dutch lipid clinic criteria, and then let me know if that sort of differs from others, because I want to make sure- That, 'cause there are different criteria for this, correct? And I think one could argue, maybe you're biased, but I might share your bias, that the Dutch criteria might be the most rigorous. Uh, is that-

   2. JKJohn Kastelein30:36

      There is the Simon Broome criteria, there is the WHO criteria, and the Dutch Lipid Clinic. Now, the Dutch Lipid Clinic criteria were put together by one of my coworkers, Peter Lansberg, and this set of criteria was externally and internally validated with mutations and huge numbers. And every time you do a comparison between the Dutch and the rest, the Dutch are winning in terms of their power to predict FH.

   3. PAPeter Attia31:05

      So it's not unlike Max Verstappen winning the most races also being Dutch, right? Just to clarify that.

   4. JKJohn Kastelein31:12

      [laughs] Well-

   5. PAPeter Attia31:13

      I don't, I don't know, I don't know how much of a fan you are of Max

   6. JKJohn Kastelein31:15

      I think Max Verstappen is actually totally brilliant guy and a fantastic athlete. This is just something that you get points. So if you have a first-degree relative with known premature coronary disease, you have a score of one. Or if you have a first-degree relative with high LDL, you get one. Then if you have children or you have these xanthomata, you get two. I know this is an audio, but it's a long list.

   7. PAPeter Attia31:41

      Mm-hmm.

   8. JKJohn Kastelein31:42

      And what's interesting is that if you have a mutation, you actually get eight, and a diagnosis of definite FH is above eight. So you have definite FH, probable FH, possible FH, and unlikely FH, which is great because you can divide your patient population into these categories, and as you're saying, it has a therapeutic consequence. Because if someone has definite FH, we treat from the age of six. That's in our national guidelines, and then we start treating immediately, just to give you a feel for that.

   9. PAPeter Attia32:19

      Let's just make sure, uh [laughs] , people understand what was just said there, right? If we can, through this very rigorous diagnostic criteria that, just for the record, was not developed by Max Verstappen, if we can establish that a person has FH with such a high degree of certainty that we would call it definite, we would be treating an individual as young as six years of age. Which means our certainty with which this person's life is at risk in as early as the third decade of their life, i.e., in their 20s, we're so sure of that that we would do something that honestly I think a lot of people who don't, especially if you don't really understand the pathophysiology of lipid metabolism, would think that's absolutely insane. Now, I obviously share your view, which is that's absolutely not insane. That's the only way that person's going to go on to live a long life. But I, I would imagine that that's a very difficult discussion to have with the parents.

   10. JKJohn Kastelein33:18

       Well, sometimes.

   11. PAPeter Attia33:20

       Yeah, yeah. I was just about to say, it would depend on what they've already experienced.

   12. JKJohn Kastelein33:23

       Absolutely. It's all basically determined by family history. I have seen in the pediatric lipid clinic, especially in the first five to 10 years, mothers with no father anymore.

   13. PAPeter Attia33:38

       Mm.

   14. JKJohn Kastelein33:38

       So mothers in their 30s, uh, their husband was deceased. She came with three children, and then one child had a total cholesterol of 10, which is like 400 or something. The middle child had normal cholesterol, and the youngest child had a cholesterol of 300 or something. And so that mother really wants her child treated.

   15. PAPeter Attia34:03

       Yeah, she knows that two of her three children inherited the genes that killed her husband at 30.

   16. JKJohn Kastelein34:10

       Exactly. Totally unannounced. The first heart attack was the last. Terrible.

10. 34:16 – 41:08

    Who ‘escapes’ FH complications and why: the rare resistant phenotype

    1. PAPeter Attia34:16

       One more question actually, John, before we get to treatment. What is your best guess as to two things? One, the fraction of people with FH who do not go on to develop premature ASCVD. So we'll call this the fraction of people with FH that seem immune to the phenotype. And then of course, the more interesting question is what would be some plausible explanations you would offer for that? So let's start with the easier question if you even, if it's knowable.

    2. JKJohn Kastelein34:51

       We estimate based on our long-term follow-up of that large Dutch cohort that about 5% of people escape any disease symptoms at all. So that includes coronary artery calcium scoring, CT angios, xanthelasmata, xanthomas, arcus, anything. They seem completely immune to the elevated LDL cholesterol. Now, I have to say that the vast majority of those are women.

    3. PAPeter Attia35:21

       Mm-hmm.

    4. JKJohn Kastelein35:22

       So that's the first thing. In men, it is extremely rare to have that. So the majority are women, and very often we discover those women when we go into the family. A child is referred, total cholesterol 380. There's no thyroid problem, no renal problem. Then we go to the father and the mother and the uncles, and then we find an aunt who's 76, has never been treated, and we measure cholesterol, and she also has an LDL of 250. And then we do mutations because in Holland the mutation screening is free. The government pays for that. So we do mutations in everyone that comes to our clinic basically free. And so then we find a mutation in her too, and there are some explanations. Very frequently, those women have very high HDL cholesterol. So perhaps some people have such an efficient reverse cholesterol transport system that they can take care of the deposited LDL in the macrophage, on the tendons, everywhere. It's almost like the reverse cholesterol transport back to the liver is so fast that you can dump anything on them, doesn't matter

    5. PAPeter Attia36:36

       So in other words, they have something protective

    6. JKJohn Kastelein36:39

       Yes. They absolutely have something protective, and so they are almost never smokers. So if you're a man and you smoke and you have FH, it's a death sentence if you're not treated. In the old days, or people that are not discovered right now, then really that interaction between one pack or, or two packs a day with LDL is the worst in terms of heart disease risk. So the women are very rarely smoking. They also almost never have diabetes, so they're thin, active women, have high HDLs and don't smoke. But we have desperately tried, we even had grants and everything to understand genetically if something maybe with these people, we and no one i- I think on the planet has ever found a real good biological reason as to why some people are so resistant against LDL cholesterol.

    7. PAPeter Attia37:36

       And I assume the sample size is not that large, but has anybody looked at monozygotic twins, both with FH, to see if there are differences in progression as a function of various lifestyle factors that might give us a sense of how much of that is driven by divergence in behavior?

    8. JKJohn Kastelein37:54

       It's very interesting, by the way. There is one of the largest monozygotic twin cohorts is in Amsterdam at the other university. Doret Boomsma, a wonderful researcher, but FH is just too rare to have enough of those monozygotic twins to make any conclusion.

    9. PAPeter Attia38:12

       By the way, I'll just add something based on the discussion with Dan Rader that you referenced, that is the HDL cholesterol story is a pretty complicated story. It's a far more complicated story than the LDL cholesterol story. The LDL cholesterol story is actually relatively straightforward. We don't know when we measure an elevated HDL cholesterol if it is a biomarker of something good that is happening, i.e., HDLs delipidating foam cells, macrophages, things like that, or if it's a bad thing. I went really down the rabbit hole three years ago on the literature of people with elevated HDL cholesterol who were developing premature ASCVD. So it posited that these people with elevated HDLC, that was actually a biomarker for very dysfunctional HDLs.

    10. JKJohn Kastelein39:02

        Yeah. Those are people with SRB1 mutations.

    11. PAPeter Attia39:04

        Exactly.

    12. JKJohn Kastelein39:05

        We found one family and published that in the New England. Very interesting. These people had very high HDL, but they had premature coronary disease. But remember, the woman I'm describing is a nice, thin, elderly grandmother who's very active, jumping up and down, doesn't smoke, and her high HDL is definitely not a sign of something dysfunctional.

    13. PAPeter Attia39:30

        My point being is this is such a complicated, for lack of a, a, a more productive word, it's a multifaceted problem for which there might just be in these 5% of people, again, mostly healthy women who seem immune to the phenotype, there's an alignment of the stars where enough other things are working in their favor that it's offsetting this damage. So let's now talk about treatment, and I want to talk about it through the lens of are there any differences in how we treat people by sex, by age, and by level of disease at time of diagnosis? So whether we're dealing with primary prevention, and that's a term, I don't know if you would agree with me on this, but I describe primary prevention as treatment when there is not a single discernible sign of disease. Whereas I would call secondary prevention, if you have a calcium score of five, we're already in secondary prevention even though you haven't had an event.

    14. JKJohn Kastelein40:31

        Or if there are abnormalities on a CTA, of course.

    15. PAPeter Attia40:34

        That's right.

    16. JKJohn Kastelein40:35

        No, I completely agree with you. Primary prevention is rare. [laughs]

    17. PAPeter Attia40:38

        That's exactly right. So secondary prevention is, yeah, your CTA might be zero calcium score, but there's a soft plaque well inside the artery. Zero luminal obstruction, doesn't matter. You have ASCVD that can be documented by the naked eye effectively, not even microscopically, so we're in the secondary prevention world. So maybe walk us through primary versus secondary prevention, men versus women, children versus adults, and take that in any order that you like as far as how you think about treatment.

11. 41:08 – 49:16

    Treating FH across the lifespan: children vs adults, targets, and modern combination therapy

    1. JKJohn Kastelein41:08

       Let's start with the child. So the child comes to us at the age of six years is the first time a child understands what they eat, and in heterozygous FH, it's very important to start with a very healthy lifestyle early on. So we give them extensive anti-smoking training, and we give them extensive dietary counseling for all the good healthy choices and also tell them sports, physical exercise, so there's a large... Okay. But that never detracts from the fact that we will start with a statin at the age of six.

    2. PAPeter Attia41:46

       Before you get to that, can you give some insight or color around what type of dietary advice? Are you recommending saturated fat restriction? Because obviously saturated fats, I think in FH you probably have a greater likelihood of exacerbating the condition. Is it true that people with FH are more sensitive to dietary saturated fat, or does it just seem that way anecdotally to me?

    3. JKJohn Kastelein42:10

       Yeah. That's, I think, anecdotally. The problem is, is that we all know that these lifestyle measures are not gonna cure FH. They are just supportive, which is very different than for the general population of course. But heterozygous FH is driven by the LDL and how long a patient is exposed to that LDL. So the earlier we can intervene, the better, though actually it's pretty clear that if you start intervening early in LDL, you probably add 15 to 20 years to that life. Versus doing nothing on the medication front. So you need to start treating the child, and of course, you start with a statin.

    4. PAPeter Attia42:50

       Do you have a preference for statin, John?

    5. JKJohn Kastelein42:52

       We used to have a preference for pravastatin because it was kind of seen as the mildest, and we were only allowed in the beginning to use pravastatin because we had done a two-year trial where we showed that carotid IMT in these children in a randomized trial of pravastatin versus placebo, there was attenuation of progression. So we showed already in kids between eight and 18 years of age, that was a JAMA paper, actually that we could stop the progression of athero when we treated them with pravastatin, and this was way back. And now there are guidelines for the American College of Pediatrics, there are European guidelines, and in some countries they start at the age of eight, other countries it's six, but it's somewhere between six and eight. And then you have statin, and I don't think there's much of a preference anymore for a certain statin. Actually, some people have a preference for rosuvastatin because you can dose it at two and a half milligram to start with, which is a tiny pill for a kid. And then you add ezetimibe because you don't go to goals like you do in adults, but you try at least to have an LDL below 130. So 130 is kind of the num- I mean, there's no, it's very interesting because there's no intervention evidence for this at all like in adults. Uh, you must have seen the, uh, FOURIER-OLE results that came just out two weeks ago where [laughs] even LDLs below 20 are better than below 55, are better than below 70. We don't have those data for kids, so-

    6. PAPeter Attia44:26

       But I was gonna say, I was actually kind of surprised to hear that you would have such a modest goal of 130 given that kids without FH have LDL cholesterol levels of 20 and 30. In other words, we know-

    7. JKJohn Kastelein44:40

       But that's fighting conservatism.

    8. PAPeter Attia44:42

       Yes, yes, yes. No, I understand. I just want to make-

    9. JKJohn Kastelein44:43

       So I don't agree. I personally don't agree with the 130, like probably you don't agree, because you and I know what the healthy LDL is for your, for your endothelium, and it's much lower than 130. So, but this is always a fight, you know, where there, there are people that say, "Yeah, maybe." We've done a lot of research. We've looked at growth, mental state, at learning ability. We've looked at maturation, hormones, menarche, puberty development, and we've never found any negative effect in kids of statins, but there are conservative people, and so we have more modest goals. But I think it's very diverse. There are pediatricians who treat kids more aggressively.

    10. PAPeter Attia45:28

        I guess my point is it's so ironic because we have the natural experiment right in front of us, which is kids are born with an LDL cholesterol of 10 milligrams per deciliter, and as they go through puberty, it starts to go up. But during the most important periods of development, i.e. when their brain is developing, it's doing so in an environment where cholesterol is virtually undetectable.

    11. JKJohn Kastelein45:55

        Absolutely. The first year, your brain grows by far the fastest, and you have basically no LDL left to go anywhere. So it's ab- completely scientifically agree with you.

    12. PAPeter Attia46:07

        And by the way, John, has the thinking changed with PCSK9 inhibitors? Because I guess you could make the argument, well, statins are impairing cholesterol synthesis, and maybe in a developing child, that could be problematic. But a PCSK9 inhibitor has no bearing on cholesterol synthesis. It's simply amplifying LDL clearance. So does that change the thinking at all?

    13. JKJohn Kastelein46:35

        Well, it changed my thinking, and I agree with you, but there are other indications, all preclinical, so pretty useless in my opinion, that in the embryonic stage, PCSK9 has a role in brain development, and there are some Mendelian randomization studies that pick up an increased signal for Alzheimer's disease with low PCSK9 while all the trials actually have never shown any effect on cognition. But of course, those arguments are again used by people that are more conservative. Listen, we have done all the trials because we have so many kids with heterozygous FH for the files of both alirocumab, evolocumab. We've done all the trials in children for heterozygous FH, also for simvastatin plus ezetimibe, all the statins alone. We've done prava, we've done rosuvastatin, we've done simvastatin, and then simvastatin and ezetimibe, and then the PCSK9 monoclonals. And so we've done all that, and the kids are... It's so interesting. Kids are not statin intolerant.

    14. PAPeter Attia47:43

        Meaning you don't see kids developing myalgias and statin-

    15. JKJohn Kastelein47:48

        No

    16. PAPeter Attia47:48

        ... related muscle symptoms-

    17. JKJohn Kastelein47:50

        No

    18. PAPeter Attia47:51

        ... at the freq- 'cause in adults, we see that about 5% of the time. You're saying-

    19. JKJohn Kastelein47:54

        Yeah

    20. PAPeter Attia47:54

        ... you don't see that in kids?

    21. JKJohn Kastelein47:57

        No. Because they don't write the inserts.

    22. PAPeter Attia48:00

        [laughs] They don't read the inserts, yeah, exactly.

    23. JKJohn Kastelein48:02

        Sorry, they don't read the inserts.

    24. PAPeter Attia48:04

        But, but do you think it's also because you're disproportionately using pravastatin, which is milder, or do you think it's true even in the presence of rosuvastatin?

    25. JKJohn Kastelein48:12

        No, it's also true in the presence of atorvastatin and rosuvastatin, yeah. And listen, it's very interesting because we have now a follow-up, and I think that was a New England paper. We have, of course, the kids that were eight are now 18, 10-year follow-up, and the kids that were 18 are now 28. So we have compared these kids with the generation above them at that age, and none of these kids Actually got a heart attack or angina or anything. And the generation above them who didn't get any treatment, of course, because there were no statins or nothing, their early mortality, early coronary disease was rampant. So I think we've already shown, although it's not a real randomized controlled clinical trial, but it's the only way you can do this, you can't randomize a kid, of course, you can only do observational data, is that treating from an early age actually protects the kids against premature death and premature coronary disease.

12. 49:16 – 52:46

    Severe heterozygous and homozygous FH: when triple therapy isn’t enough and the role of apheresis/evinacumab

    1. PAPeter Attia49:16

       Okay. So I wanna move on to another topic, but before I do, I guess the last thing I wanna contrast this with is the adult, middle-aged or otherwise, who shows up either with or without disease. How aggressive do you go there?

    2. JKJohn Kastelein49:34

       As aggressive as the guidelines tell us to go for non-FH patients. So the kids are transferred from the pediatric lipid clinic to the adult lipid clinic when they're 18, and then we immediately start with PCSK9 monoclonals and pretty soon with inclisiran and the ezetimibe, and we really strive for the lowest LDL possible for that individual. And in the homozygotes, we do high-dose statin, ezetimibe, evolocumab, and now evinacumab, the Regeneron NPTL3 monoclonal antibody. So those four things constitute the state-of-the-art therapy for homozygous FH. It gets a lot of patients to relatively normal LDL levels, and that is a miracle for homozygous FH. In heterozygous FH, there are about 10% of patients that even with triple therapy, we can't get them to a reasonable LDL. So we call these now severe heterozygous FH. They probably have more than one mutation. They probably have another mutation somewhere else that makes it a more severe phenotype, but that, we officially don't know that.

    3. PAPeter Attia50:53

       Meaning these patients on PCSK9 inhibitor plus statin plus ezetimibe, you can't get below 70 milligrams per deciliter? What-

    4. JKJohn Kastelein51:03

       Oh, no. We don't get them below 100 or below, uh, 120. So some of them are, are really, they've nasty opposition against your therapy. And it's linked, by the way, to the starting LDL. So we say that if your starting LDL is above 300, you have severe heterozygous FH, and that's about 10% of the overall heterozygous FH population. Of course, when you start at 300 and you have a statin that takes off 45%, then another new baseline, 15 to 20, and then another, then you can calculate that it's not that easy to get below 100.

    5. PAPeter Attia51:42

       Is apheresis even a viable option? Is it easy to pull out? I know it reasonably works for Lp (a) , but does it work for the majority of the ApoB-bearing lipoproteins?

    6. JKJohn Kastelein51:52

       It does. We have an apheresis center in Amsterdam because we've sometimes diagnosed homozygous FH after six months in kids of six months that had a very severe homozygous FH, and then we've had two LDL apheresis kids. But I can tell you one thing, evinacumab is the golden rescue for these children. It's still only, uh, used in adults, but I hope that we will be able to use that quadruple therapy rapidly in kids also because it obviates the need for the apheresis in many instances.

    7. PAPeter Attia52:31

       What's the frequency that it needs to be given? The drug is dosed how frequently?

    8. JKJohn Kastelein52:36

       It's intravenous dosing, I think, once a week.

    9. PAPeter Attia52:39

       Wow.

    10. JKJohn Kastelein52:39

        But they are working on a subcutaneous formulation of it because, of course, intravenous with, you know, these kids are, are used to something.

13. 52:46 – 55:47

    LDL causality debates: FH as a natural experiment and why exceptions don’t refute the rule

    1. PAPeter Attia52:46

       Yeah. I mean, these kids don't have veins. I mean, you're basically putting PICC lines in kids to give them treatment or something. Yeah. Wow. So there is a small but vocal cadre of people out there who kind of refuse to believe that LDL is causally related to ASCVD, or another subvariant of this group who believe that it's only causally related to ASCVD in the context of metabolic illness, but if you're metabolically healthy, then LDL is not problematic. Based on your knowledge and experience with FH, which spans the spectrum of metabolically healthy to unhealthy people, how likely do you think that is?

    2. JKJohn Kastelein53:32

       These patients that you mentioned before are always used as the stick to beat people like me because they say if LDL is truly causal, it's impossible that some people don't get heart disease while having FH. I try to explain that most dis- genetic diseases are modified by other environmental and genetic factors, but someone, these cholesterolcritics.org, it's a website of, of a strange bunch of people that absolutely don't wanna accept that LDL cholesterol is bad. But my experience with FH, my entire career, I've seen these families, I've worked 100-hour weeks for years to find all these people, and we were able to find the largest cohort on the planet of heterozygous FH, and these stories are all heartbreaking. So for me, it is so simple because these people have one thing, one, a mutation in a single gene that doesn't do anything else than raise LDL cholesterol, and they drop dead, you know, when they're, like, 25. So for me, I don't even listen to it anymore.

    3. PAPeter Attia54:44

       Yeah. I mean, it's the sort of Mendelian randomization also makes that clear because it doesn't just include FH, it goes all the way down And my response to people when they point out the observation that there are some people with high LDL cholesterol or high ApoB who don't have coronary artery disease is, there are also people who smoke their whole lives and don't get lung cancer. And by the way, there are people who never smoke who do get lung cancer. Neither of those facts remotely diminishes the causal case for smoking and lung cancer.

    4. JKJohn Kastelein55:17

       I couldn't agree more, Peter. Absolutely. And I find it also, because of course, the FH argument was always the scientific argument of proving that LDL cholesterol, if it's elevated, sits at the core of atherogenesis. And then, of course, when it became clear that there were a few people among these families who could actually live without a problem, they used that as an argument. But I completely agree. The smoking is a very good example, but there are many more.

14. 55:47 – 1:13:04

    CETP inhibitors: biology, evolutionary framing, and the ‘sordid history’ of drug failures

    1. PAPeter Attia55:47

       So let's pivot now to a class of drugs that I discussed at length with Dan Rader, the CETP inhibitors, because you're very involved in not just the latest of these, but potentially the first one that really appears to have a shot at working. So let's back up for a second, and for folks who either don't remember the podcast with Dan or didn't hear it, let's provide an explanation of what CETP is and maybe a little bit of the sordid past of the CETP inhibitors and, you know, basically what is prologue to where you are now.

    2. JKJohn Kastelein56:29

       The prologue, unfortunately, is very long, Peter, [chuckles] and it's one of the best examples of how big pharma can make big mistakes. So the CETP protein was discovered by Phil Barter, the Australian KOL that's now retired, but he discovered this protein, and I think he actually discovered it in rabbits. And why did he look in rabbits? Because if you give a rabbit egg yolk or cholesterol, that rabbit gets atherosclerosis. If you do the same diet to a mouse or to a rat, or a hamster for that matter, they don't get athero. And so it was extremely interesting what made a rabbit different from other rodents, and that is CETP. So a rabbit has CETP, so all strategies to lower CETP were tested in that New Zealand white rabbit, like siRNA, gene therapy, small molecules, antibodies against CETP. And in all instances, you could cure the athero with lowering of CETP activity. At the same time, it became obvious that in Japan, there are many people with a mutation in CETP, so they have very low CETP. And initially, the first report said these people live longer than we do, and they are free of coronary disease. So then loss of function of CETP became a longevity gene. It is still by many people called a longevity gene, because if you don't have it, you live longer, you have less Alzheimer, you have less diabetes, and in general, you are just simply more healthy.

    3. PAPeter Attia58:16

       Sort of like PCSK9 loss of function as well, also a longevity gene.

    4. JKJohn Kastelein58:20

       Yes. You don't need that. You don't need PCSK9. The theory is, is that we all went through the evolutionary funnel about 10,000 years ago during the last ice age. There were very few humans alive, especially in the north, and they went through the evolutionary funnel where everything was directed against being thrifty. People that could absorb the last calorie out of a mammoth were favored, and actually, we now think that those genes that we selected during that evolutionary funnel are now bad for us. Like, we don't need PCSK9. We don't need CETP. We probably also don't need NPTL3. And so all of these genes were meant to conserve energy and to conserve cholesterol. And so what CETP does is a very simple thing, is it grabs a cholesterol ester molecule from HDL, from an HDL particle. CETP sits on the HDL particle like a little cap. It's a curved protein that sits on top of a sphere, and it has an opening, and it sucks a cholesterol molecule out of HDL. Then that particle collides with an LDL particle, and there's a little tunnel, and it spits the cholesterol ester molecule straight into LDL. So the consequence of that is, is that HDL cholesterol goes down and LDL cholesterol goes up, which in the days of very active LDL receptor activity was a great idea because then all the cholesterol went back to the liver, and a cholesterol molecule is very expensive to make. It costs 27 ATPs. In a nasty environment like the ice age, you wanna conserve that molecule, and the best way of conserving a cholesterol is by sending it back to the liver, and then the liver can decide what to do with it, put it in VLDL or in bile or, or it can do a lot of things with it. But in a situation where our LDL receptors are not that active anymore, adding cholesterol to LDL is not a very good idea. And all the Mendelian randomization studies have shown that people with high activity of CETP have more heart disease, more heart failure, more kidney disease, more diabetes, more Alzheimer, blah, blah, blah, the whole works. And so Pfizer was the first to say, "Ah- Atorvastatin's patent is going to expire within a few years. Let's have another atorvastatin. So we make a CETP inhibitor called torcetrapib. I was involved in all these large trials, so I was on the steering committee of the evacetrapib trial, on all of these trials, except REVEAL. That was done by Oxford. But Pfizer did a phase two, and in the phase two, after four weeks, blood pressure went up a little bit, and then comes the big pharma mistake. Why do we care about a little blood pressure increase if HDL goes up by 70%? They didn't understand why the blood pressure went up, and they moved the drug into phase three. That is a fundamental mistake. If you don't understand a side effect in phase two, you don't move a drug into phase three until you've understood it. But listen, it was my fault also. I was on the executive committee of the outcome trial with torcetrapib, and we all thought, "Oh my God, this is gonna... You know, this is gonna be the next fantastic wave of drugs in cardiovascular disease." But that drug killed more people than it saved.

    5. PAPeter Attia1:02:05

       I remember in September of 2006 exactly the street I was standing on in the financial district of San Francisco when the trial was announced that it was being halted, and I was so s- I mean, to quote, uh, Planes, Trains and Automobiles, if I had woken up with my head sewn in the carpet, I would've been less surprised than that outcome because I, too, was so optimistic based on the HDL cholesterol increase, even though it was clear this trial was kind of a stupid land grab from an IP perspective on the part of Pfizer. You know, you had to combine it with atorvastatin and all that kind of stuff.

    6. JKJohn Kastelein1:02:47

       Exactly. Exactly. Yeah.

    7. PAPeter Attia1:02:48

       Putting that aside, I thought this is gonna be a world beater, and it turned out to be an enormous failure.

    8. JKJohn Kastelein1:02:54

       Then everybody was fired except the basic scientists who got the assignment to understand what happened, and they... It was so interesting because they took the drug, and they infused it into a rat. Now, a rat does not have CETP, but the blood pressure went up in 10 minutes. Boof. So that told everyone that this is a drug that has an off-target effect, fortunately has nothing to do with CETP, but the drug actually raced straight into your adrenals, where it promoted aldosterone production, cortisol production. It raced into your endothelial cells, where it promoted endothelin-1, which is like an angiotensin II, terrible vasoconstrictor, and all of that led to water and sodium retention, low potassium, high blood pressure because these were secondary prevention patients, and you don't want a drug like that in a secondary prevention patient. So that was the most unfortunate beginning of a new story in our field, the wrong drug.

    9. PAPeter Attia1:04:01

       You know, it's so funny. This was not too long after another epic failure, but one that I would argue resulted in a drug being removed that shouldn't have been removed, but instead required a little bit more work to determine who the susceptible individuals were, and of course, that's the drug Vioxx. I don't know if it had the same name in Europe, but truly a remarkable COX-2 inhibitor, in a league of its own, makes Celebrex look like drinking water in terms of its impotence. And there's no question that there were a subset of patients in whom Vioxx slightly raised blood pressure and led to a small increase in events. But I think Merck, I think it was Merck that was the company that made Vioxx, their... Again, I, I don't wanna over speak, but I think their arrogance and refusal to act in a timely manner resulted in just the loss of, of a drug that I think to this day we're- we'd be better off with than without. There's just something about big pharma where they are often deserving of the reputation they have.

    10. JKJohn Kastelein1:05:01

        [laughs]

    11. PAPeter Attia1:05:01

        Not always-

    12. JKJohn Kastelein1:05:02

        Yeah

    13. PAPeter Attia1:05:02

        ... but often they are their own worst enemies.

    14. JKJohn Kastelein1:05:06

        Yeah. This was a prime example of how can you make such a... I mean, I, I'm now a drug developer myself, so with Michael Davidson, whom I greatly admire, the idea that we would push a drug in phase three while not understanding a side effect in phase two is just incomprehensible. And then what's interesting, and not many people realize that, Peter, then came Roche.

    15. PAPeter Attia1:05:30

        Yes, and this is what? This was, uh, dalsa-

    16. JKJohn Kastelein1:05:32

        Dalcetrapib

    17. PAPeter Attia1:05:33

        ... cetrapib, yeah.

    18. JKJohn Kastelein1:05:35

        Dalcetrapib. That drug was extremely important for all the science because that drug only raised HDL. It was a very weak CETP inhibitor. It only raised HDL by 30%, and there was no effect on the Kaplan-Meier curve in the cardiovascular outcome trial whatsoever. And what people didn't realize, that drug was the end of the HDL hypothesis because there was no effect on LDL, no effect on ApoB, no effect on non-HDL, but there was a 35% raising of HDL cholesterol, and that did not translate into one less heart attack or stroke. So that ended the HDL hypothesis in a way.

    19. PAPeter Attia1:06:17

        Yeah, and I think we would only go on afterwards to see two Mendelian randomizations, one looking at genes that raised HDL cholesterol, one looking at genes that lowered HDL cholesterol. Neither were found to be causally linked to ASCVD. So when I hear people tell me that their high HDL cholesterol is protecting them from coronary artery disease in the presence of high LDL cholesterol, I have to restrain myself in the context-

    20. JKJohn Kastelein1:06:46

        [laughs]

    21. PAPeter Attia1:06:46

        ... of CETP inhibitor failure number 1, 2, 3, 4, 5, 6, 7 and Mendelian randomizations that, as you said, completely fly in the face of this hypothetical belief system.

    22. JKJohn Kastelein1:06:59

        But what is extremely interesting is that in those days, we didn't understand that you really need to lower LDL cholesterol with a CETP inhibitor to see an effect on ASCVD. So Merck then actually was the third. They got a drug called anacetrapib, and by that time, people weren't sure anymore whether they had to power an outcome trial on the basis of the HDL cholesterol increase or the LDL. And Merck said to Oxford, "Just make the trial large enough," and they did a 30,000-patient trial, which is, I think, until now, still the largest cardiovascular outcome trial ever. 30,000 patients, but they overestimated the LDL lowering. The drug only lowered LDL by 17%. The baseline LDL in that trial was 60 milligram per deciliter, so the absolute LDL was 11 milligram per deciliter. That predicts, if you put it on the meta regression line, that predicted a 9% reduction in MACE, and that's exactly what they got. So that trial validated that CETP inhibition only lowers heart attacks by virtue of its LDL lowering, and that it answers to the same law as statins, ezetimibe, and PCSK9 monoclonals. It is on the same regression line. So then Michael and I understood that what we needed to find was a CETP inhibitor that didn't have the off-target effect of torcetrapib, that was way more potent than dalcetrapib, and lowered LDL robustly so we could repeat the anacetrapib trial, but then with a much bigger effect size. And we found that drug at Mitsubishi.

    23. PAPeter Attia1:08:53

        Before we talk about obicetrapib, I want to kind of highlight two things you said. The first is you have literally provided the most lucid evolutionary explanation for our species-wide transition to the preservation of ApoB. I have to be honest with you, I had never heard it explained the way you did, and it makes so much more sense than any other kind of teleologic or evolutionary explanation. So I just want to make sure I heard it correctly, because I'm going to use it often. I'm going to be the most boring guy at the parties now, because I'm going to kick s- u-use this story to explain it. [laughs]

    24. JKJohn Kastelein1:09:35

        It's called the thrifty gene hypothesis, and it's more often used to explain, for example, that people in Asia get type 2 diabetes at a much lower BMI than we Caucasians, because they've gone through much more famine. When the rice failed, huge famine in the Far East.

    25. PAPeter Attia1:09:58

        Yeah, yeah. And I've always been familiar with those arguments as they pertain to diabetes and obesity. I just had never taken the additional leap of, hey, I know it takes 27 ATP to make a molecule of cholesterol, but I never made the additional leap, which was think about how expensive that is in, in an environment that is so resource-constrained, which up until 150 years ago, we were. So we spent hundreds of millions of years in an environment where preservation of resources was the second priority only after reproduction, that of course we would be so effective at LDL clearance and therefore, of course, we would want to be in the business of shuttling as much cholesterol from HDL into LDL via RCT because we knew it was going to a good place. Back to the liver, it would be circulated as bile. We would ultimately recirculate that pool. We could make more hormones. We could digest more food stuff. Yeah, it all makes sense. And then lo and behold, uh, about effective like 150 years ago, there's not such a premium on that now, and we've got more than enough cholesterol to spare, and that thing, that gene that we worked for millions of years to preserve now is biting us in the ass, just like the same genes are around adiposity and insulin resistance.

    26. JKJohn Kastelein1:11:28

        And there is a very interesting additional piece of evidence, is that if you make a mouse look like a human heterozygous FH, and you infect that mouse with bacteria and compare the results of infection with bacteria in the non-FH mouse, the FH mouse is better resistant against bacterial infection. And so the theory is, and we've seen that in our pedigrees, is that FH in 1860 was not such a severe disease as it's right now, and maybe there is so much FH because, again, in the Ice Age, having high LDL might have been an advantage as a protection against bacterial infection. These two things converge.

    27. PAPeter Attia1:12:18

        Yeah, and think about how amazing that is in terms of a parallel to ApoE4.

    28. JKJohn Kastelein1:12:22

        Exactly.

    29. PAPeter Attia1:12:23

        Like this is the exact same story as ApoE4, which basically was the only ApoE isoform we had until, what, 200,000 years ago?

    30. JKJohn Kastelein1:12:32

        Yeah, something like that, I think, yeah.
15. 1:13:04 – 1:15:01

    Obicetrapib: discovery at Mitsubishi, potency, phase 2 results, and the phase 3 program

    1. PAPeter Attia1:13:04

       let's recap where we are with obicetrapib, right? That's the, that's the current one.

    2. JKJohn Kastelein1:13:09

       Yes.

    3. PAPeter Attia1:13:10

       So it is how potent? I guess you've done phase two-

    4. JKJohn Kastelein1:13:13

       Yeah, I don't... Listen, I, I, I am still a scientist, and I don't wanna sound like a salesman, but we were lucky finding this drug in Mitsubishi.

    5. PAPeter Attia1:13:23

       Yeah, so tell us more what that means. So help folks understand how do you find a drug? How is drug discovery done? How, how did this thing come about?

    6. JKJohn Kastelein1:13:30

       So I am a consultant to many biotech companies, and sometimes companies ask me to look at the pipeline and see if there's anything good or bad in it. So I was invited by Mitsubishi to look at a number of things, and I saw the phase one data of this compound that was still called TA-8995, and I saw that at one milligram, that drug lowered LDL by 27%. Now, remember, dolsetrapib was used at 600 milligram, and it didn't do anything on LDL. So I thought, "Holy moly, this drug is very potent." And then I looked at the CETP inhibition. At 10 milligram, which is the dose we're now using, there's a 97% inhibition of CETP. So it's simply, in that sense, and surprisingly potent CETP inhibitor, and that translates into an LDL lowering of about 50% on top of high-intensity statins and an HDL increase of 165%. Now, you have to control yourself when you see those numbers because you know that the HDL cholesterol increase is not gonna do anything for heart attacks and strokes. But Michael and I are not only working on this drug to develop it, we're also scientists, and we wanted to understand all the genetic and epidemiology and Mendelian randomization data. And so

16. 1:15:01 – 1:41:41

    Unexpected metabolic signals: diabetes risk reduction, septicemia protection, and Lp(a) lowering

    1. JKJohn Kastelein1:15:01

       we've gone far beyond heart disease with this drug. We're looking at Alzheimer, age-related macular degeneration, septicemia, and diabetes. Because if you really inhibit CETP, you not only stop the transfer from cholesterol from HDL to LDL, but you completely change lipoprotein metabolism, and you force the liver to produce more ApoA1. And you and I can, I think, agree on the fact that ApoA1 is a fantastic molecule because it is the molecule that ABCA1, the cholesterol pump on the cell membrane, recognizes and actually exports cholesterol to. So that was a first. And then we also know that if you produce lots of ApoA1, you produce lots of pre-beta HDL particles, the small HDL particles, and these particles suck cholesterol out of peripheral tissues. Now, in endothelial cells and macrophages, that's probably good, but what's much more interesting is that they suck cholesterol out of the beta islet cells in your pancreas. And you undoubtedly know that with life, if you're a type 2 diabetic, more and more of these cells die till the point that you become insulin dependent. That is because of the lipotoxicity of that cell. That cell takes cholesterol, can't export it, the sterols are oxidized, they become pro-inflammatory, toxic, the cells go into apoptosis, and they die. So that is alrea- And this is proven now for all four CETP inhibitors. So all four CETP inhibitors in their outcome trial had less diabetes in the treatment arm than in the placebo arm, and we've published that in a meta-analysis about half a year ago.

    2. PAPeter Attia1:16:59

       Just to be sure I understand, John, is that because you think regardless of which of the CETP inhibitors we saw, we were seeing more heterotypic exchange between the ApoBs and the ApoNs and both? Okay, so let's explain for people. Tell them the difference between homotypic exchange and heterotypic exchange, which is pretty easy to define, but more importantly, clinically, where these are occurring and why these are leading to the outcomes that we're about to get into.

    3. JKJohn Kastelein1:17:30

       So the sequence is probably this. You stop the transfer of a cholesterol molecule going from HDL to LDL. That will make HDL higher by definition and LDL lower. Now, Dan has done stable isotope turnover studies and have shown that the reason for the lower LDL is that they are swept up by the liver. So these LDLs actually get 50% lower, which is a huge drop, because the liver upregulates, and what can the liver upregulate? Their LDL receptors. At the same time, these large HDLs take ApoE on board, and as you know, ApoE is also a ligand for the LDL receptor. So the large HDL particles are also cleared by the liver, and the liver does that because it produces large amounts of ApoA1, which kind of disturbs the balance. So there is a new equilibrium between removal of lipoproteins by the liver and production of small HDL particles, but that only happens if you almost completely knock out CETP. So you have to really inhibit it by about 90%, which is very close to the homozygous patients in Japan. They also have half LDL and about a triple HDL and every other thing that I just described. So it's For me, and, and I think also for Tom Day speaking and people that really know lipids, this is extremely interesting because it is complex. It is complex, but until now it is extremely exciting because there are two groups in the world. So this is the diabetes part, you know, where you suck cholesterol out of the beta cell. Then there is the septicemia part that I never really knew about. There are two groups in the world, one in Vancouver and one in Leiden, that have shown that if you're born with a loss-of-function variant in CETP, you are much better protected against septicemia, and not a little bit, but like mortality, big effect. Actually, there are multiple presentations this year of these two groups, and they, of course, want our drug to test it in septicemia.

    4. PAPeter Attia1:19:58

       Let me make sure I understand something, and I want to go back and ask a question about FH to bring it back to that. Is there any evidence that FH patients untreated have a lower mortality due to septicemia?

    5. JKJohn Kastelein1:20:10

       It's not great science, but there are some indications, but most of that work is done preclinically.

    6. PAPeter Attia1:20:17

       So it's not as straightforward as, look, if you have high peripheral cholesterol, you have more precursor to make corticosteroids and therefore support immune function in the time of sepsis. It's, it's much more complicated than that.

    7. JKJohn Kastelein1:20:38

       No, it's way more complicated, and it's most likely has to do with the scavenging function of lipoproteins. So if your CETP is low, when you get septicemia, your HDL stays high, and it's very likely that the HDL particle functions as a sink for endotoxins and everything else. And so having a high HDL and a constant high HDL during septicemia is a very good thing.

    8. PAPeter Attia1:21:08

       And I, on a recent podcast, shared that I used to witness the opposite in the ICU, which was a drop in HDL cholesterol.

    9. JKJohn Kastelein1:21:17

       Yes, you always witness that.

    10. PAPeter Attia1:21:18

        Yeah. And, and that would, that would actually be a poor prognostic indicator. Not, not necessarily poor, it's just that's the normal indication. Yeah, yeah.

    11. JKJohn Kastelein1:21:25

        Yes. If your, if your HDL drops like to nothing, that's a poor prognostic indicator. You are protected against that drop if your CETP activity is low because it's like having a CETP inhibitor on board which stops removing cholesterol from HDL.

    12. PAPeter Attia1:21:43

        Yeah. So again, it-- this is a function issue that is not just a, a blind phenotype issue. The sort of paint-by-numbers approach to this problem is high cholesterol, good, because it's more precursor. No, no, it's CETP inhibition, good, because you have a bigger sink to dispose of toxic waste.

    13. JKJohn Kastelein1:22:04

        Absolutely. That's it. So that is the septicemia part and the diabetes part. Now, the diabetes part is proven in a meta-analysis. For example, the ro- the Roche drug, and that's so interesting, the Roche drug only raises HDL.

    14. PAPeter Attia1:22:22

        Doesn't lower LDL.

    15. JKJohn Kastelein1:22:24

        No. Zero. And so its protective effect against diabetes can only be connected biologically with the HDL, of course. And so it's now thought that if you raise HDL with CETP inhibition, you protect the pancreas against apoptosis. It's an effect of about sixteen to twenty percent, huh, in the new onset type 2 diabetes between placebo and active treatment arm. So it's not like a tiny effect. It's almost as big as the negative effect of statins on diabetes, which is also at, at the highest dose, fifteen percent or so.

    16. PAPeter Attia1:23:01

        So let's just make sure, again, we bring it back to this idea because am I correct in saying that obicetrapib's the first CETP inhibitor that impacts both heterotypic and homotypic exchange?

    17. JKJohn Kastelein1:23:16

        That is very hard to say, uh, Peter, because there is not much work done with the older CETP inhibitors. Dan Rader has done a huge amount of work with the Merck anacetrapib. He has, for example, shown that HDL particles from anacetrapib-treated patients have more cholesterol efflux from macrophages. He has shown that SR-B1 is upregulated in the liver by CETP inhibition with that drug. So he has done a lot of that work, but with the other drugs, not so much. Not so much.

    18. PAPeter Attia1:23:51

        Okay. This is pretty exciting. A sort of student of this world would have to be forgiven for having a little bit of anxiety and fear about being too optimistic here, right? This is one of those things where we've been burned on every one of these, right? There's a track record of four or five consecutive failures, some more epic than others, perhaps none more epic than Pfizer's. And yet, let's just say like if, if you were a Wall Street analyst trying to get your arms around this, you'd have a really hard time getting excited based on how many times you've been burned. But we're not stock analysts, and we're really just trying to come at this through the lens of science. What would you say is the greatest risk that this does not pan out? So let's explain what has, what we know so far. You've completed phase one, so there's no, there's no toxicities-

    19. JKJohn Kastelein1:24:49

        And phase two

    20. PAPeter Attia1:24:49

        And phase two

    21. JKJohn Kastelein1:24:49

        We've also completed phase two.

    22. PAPeter Attia1:24:51

        And you had four phase two trials?

    23. JKJohn Kastelein1:24:53

        We have one, two, three, four phase two trials. Yes. And we are now fully in phase three, and in fact, all our phase two trials will be fully randomized this year. So we have gone-- we've pulled every plug we could. We've started the outcome trial at the same time as our lipid trials because we, of course, realize that what we need to show is robust LDL lowering. Robust non-HDL lowering, robust ApoB lowering, and then we need to show safety. So blood should be- no blood pressure effects, nothing. Good safety, good tolerability. It's a tiny pill, a 10-milligram pill. And then, of course, at the end of the day, we also need in the future to show outcomes in a cardiovascular outcomes trial, of course. But I have to say that until now, Peter, this drug is well-tolerated. We haven't seen any side effects in phase 2, and it lowers LDL by half, which makes it just as potent as the injectables, but then in a 10-milligram pill. When we first saw these LDL results on top of high-intensity statins, I was kind of amazed at this.

    24. PAPeter Attia1:26:05

        Yeah. So I don't want to get too far ahead of ourselves, but I'll just assert something here, which is if this pans out to be as good in large phase 3 studies as it has been in phase 2, the only thing that would stand in the way of this displacing every statin on the planet is cost because you'd have equal or better lipid-lowering efficacy, and instead of having a small increase in the risk of type 2 diabetes, you would be patently reducing that risk. We're gonna talk about the brain in a moment as well. There may even be some benefits there, but let's put that aside for the moment. So there's a lot riding on this.

    25. JKJohn Kastelein1:26:45

        It's cheap, Peter. It's very cheap to make.

    26. PAPeter Attia1:26:48

        Well, I was gonna say, given that it's not an injectable monoclonal antibody, this-

    27. JKJohn Kastelein1:26:51

        Yeah

    28. PAPeter Attia1:26:51

        Yeah

    29. JKJohn Kastelein1:26:51

        It's cheap to make and, um, yeah, it's actually very cheap to make. I think this is public knowledge. It will cost $36 per year to make this drug at the time when you have lots of patients, so what's called peak sales. So the drug is cheap to make, and that will allow us to price it reasonably and ethically, and that is something that Michael and I have always wanted, is to have a drug that lowers LDL that you can use on top of a statin or in a fixed-dose combination with ezetimibe that robustly lowers LDL and ApoB, has little or no side effects, and is easy to put in your pillbox. Because the vast majority of my clinic in Amsterdam, they have pillboxes, and they are a bit afraid of a needle, and they, they want an extra pill, uh, you know, to get their LDL down.

    30. PAPeter Attia1:27:45

        So let's talk about the three trials. Just to be clear, the three phase 3 trials, you have Broadway, Brooklyn, and Prevail, right? So Broadway is 2,400 patients, one year. It's looking at drug-
17. 1:41:41 – 1:51:53

    CETP, APOE4, and brain cholesterol handling: a proposed Alzheimer’s mechanism and a CSF biomarker trial

    1. JKJohn Kastelein1:41:41

       So if you look at late-onset sporadic Alzheimer, which is the vast majority of Alzheimer patients, 65% have an APOE4 molecule on board. They're either E4E4 or E3E4 or E2E4. So they have an E4. If you have E4E4, your risk for Alzheimer is 16 times higher than when you have E3E3. And if you have E3E4, your risk is about four and a half times higher. We now begin to understand why carriership of APOE4 is so bad for the brain. APOE4 is an insufficient molecule to get cholesterol out of cells in the brain that have too much or bring cholesterol to cells in the brain that have too little. So it fails on both accounts. It is not a good acceptor of cholesterol, and it's a bad bringer of cholesterol. And so if you are E4 and those cholesterol abnormalities accumulate over life, you actually get sterol accumulation in neurons, and if you have sterols in a cell too long, they get oxidized. I mean, everything oxidizes in us. We, we rust. Like anything else, if you're exposed to oxygen, we get oxysterols. Oxysterols are what kills cells. It gives a pro-inflammatory signal. It drives cells into apoptosis. It is the worst thing that you can have. Now, an oxysterol is not much of a problem if you have a functioning particle in your brain that sucks the sterols out of the cell and then converts it to 24-hydroxycholesterol, and that gets through the blood-brain barrier to your liver, up into bile, and it's gone. That whole normal process where the brain needs a lot of cholesterol for myelinization, building synapses, building these sprouts And at the same time, if they have too much, they want to get rid of it fast in order not to get... is all wrong in an APOE4 carrier. So what is the protein that can help here? There's only one protein, APOA1. APOA1 can take over all these functions of APOE4. And how do you get APOA1 in the brain? By raising it substantially in the circulation, because unlike APOE, APOA1 can get through the blood-brain barrier because it's small enough, and there's very likely a specific receptor that actually pushes it through brain cells. And which drugs do raise APOA1 by far the most? CETP inhibitors. So that is the connection. So these large HDLs, they acquire APOE because they get larger, and they lose their APOA1, and we think, and we have preclinical already some evidence for that, but we're doing a trial in humans where we tap CSF to s- to look at APOA1, that if your APOA1 concentration goes up in the circulation enough, you'll push it into your brain, and once it's in the brain, it takes over the function of this dysfunctional APOE4. It's a fantastic story, Peter.

    2. PAPeter Attia1:45:08

       The nomenclature makes this complicated, so I guess let's make sure people understand what we're talking about. It's always a problem when the gene and the protein have the exact same name, [laughs] right? At least with Lp (a) , we have the LPA gene and then Lp (a) , the lipoprotein. But here, it's the same name. We denote it differently, right? We use all caps versus not, and italics and all that. But when we talk about the APOE4 gene having these three isoforms, two, three, and four, you're gonna get six different combinations of them. But what you're talking about is the protein. You're talking about the thing that is made by the gene. Again, what's just another remarkable insight into the complexity of biology is the protein APOE, which is not designated two, three, or four, it's just the protein APOE. The one that is made, that is transcribed and translated by the E4 isoform, I believe it only differs in one amino acid from the wild type. It's a very subtle difference, right?

    3. JKJohn Kastelein1:46:10

       It is. It is an arginine for a glutamine, I think, or something at amino acid 152 or 118. I can never remember it, but-

    4. PAPeter Attia1:46:17

       I always thought it was 127, so yeah, we're... one of us is wrong, but yeah, whatever.

    5. JKJohn Kastelein1:46:21

       Yeah. [laughs] Well, I'm, I'm sure I'm wrong. It's only one amino acid, and that changes the three-dimensional conformation of that protein completely and makes it basically a lousy cholesterol acceptor and transporter. If you carry an E4, you have, like, a list this long of things that go wrong in your brain. It's pro-inflammatory. It is insufficient in lipoprotein metabolism. It is no longer a chaperone for beta amyloid. You know, it's like a very long list, and it's terrible for, for people that have E4. I think one of these Hollywood actors actually knows that he is, uh, he is an E4 carrier. He's that, that athlete.

    6. PAPeter Attia1:47:05

       Chris Hemsworth. Yeah, he disclosed this during the, uh, the Limitless series, which, which I, I think was, was really valuable for a lot of people to see to bring a lot of awareness to this. And by the way, I've made the point many times that it is very valuable for someone like Chris to know that he has APOE4, because the earlier you take steps to prevent the exacerbating risk factors, the better your odds. I mean, the one thing that's important to point out, you, you opened this discussion by explaining the risk factors. I, I look at data that suggests it's a little less than 16 and fourfold. I think it's sort of maybe closer to tenfold and twofold, but we don't need to worry about that. There's no question it's high risk, but it's important to note it's not deterministic.

    7. JKJohn Kastelein1:47:49

       No, it's not deterministic.

    8. PAPeter Attia1:47:51

       And in fact, it's far less penetrant than FH is.

    9. JKJohn Kastelein1:47:55

       Yes, it is.

    10. PAPeter Attia1:47:55

        There are lots of people walking around with APOE4 that are not getting Alzheimer's disease, and as you pointed out, a third of people with Alzheimer's disease don't have APOE4. So everything we're talking about here is, is fair game to everybody. What you said that's interesting that I didn't realize until today was APOA1 can traverse the blood-brain barrier, and therefore, if you have a therapy that raises APOA1, you can potentially offset the damage of a defective APOE in response to APOE4. Do we have a sense from any of the preclinical work that's been done or even the early clinical work that's been done what the magnitude of that can look like? Where I'm going with this, of course, is how could we begin to quantify the potential benefit of this? Is this something where... As you probably know, Mike Davidson, also very involved in the klotho space, right? Also very involved in looking at the observation that those with klotho KLVS variants seem to have almost complete protection from their APOE4 gene. I mean, it's a remarkable finding.

    11. JKJohn Kastelein1:49:09

        Yeah. So there is one large Mendelian randomization study in E4 carriers, and which gene can protect you? Loss of function of CETP. So we already have Mendelian randomization data in humans that low CETP protects an APOE4 carrier against Alzheimer's. But what you just asked me is, can we have an effect size here? We're doing a proof of concept trial where we tap cerebrospinal fluid, and we look at, I mean, like 50 biomarkers, because what we hope is that APOA1 goes up in the brain. That the consequence of that is that the cells are gonna normally synthesize cholesterol, so the desmosterol and lathosterol levels should go up because cholesterol synthesis is normalized again. At the same time, 24-hydroxycholesterol should also go up because cells are getting rid, rid of cholesterol in a normal fashion, and the inflammation biomarker should go down because you basically substitute for this dysfunctional E4. So in order to understand the effect size, we need to be able to make a story where we say dysfunctional E4 replaced by A1, normal cholesterol synthesis is on again, normal cholesterol removal is on again, inflammation goes down, and of course this proof of concept trial is only six months, so there won't be much in Alzheimer biomarkers. But if we can show that we improve lipoprotein metabolism in the brain, it is a first step into a fascinating journey, I would say.

    12. PAPeter Attia1:50:55

        Will you be measuring desmosterol and lathosterol in any of these trials or in all of them?

    13. JKJohn Kastelein1:51:00

        Yes. Yes. No, we, we, uh, in the Alzheimer trial we do. Yeah.

    14. PAPeter Attia1:51:04

        Which is PREVAIL is the one where you're gonna-

    15. JKJohn Kastelein1:51:06

        No, no, the Alzheimer trial has no name. The Alzheimer trial is a trial in-

    16. PAPeter Attia1:51:11

        Oh, it's a fourth tri- it's a fourth trial that-

    17. JKJohn Kastelein1:51:13

        It's a fourth trial. Yes, it's a fourth trial.

    18. PAPeter Attia1:51:15

        I see.

    19. JKJohn Kastelein1:51:15

        But we have no name for it because it's in a single center in Amsterdam, and it's a large Alzheimer's center, where we have basically mild cognitive impairment with a diagnosis of Alzheimer's, so early Alzheimer. We give them our drug, and we serially measure both blood and CSF by spinal taps.

    20. PAPeter Attia1:51:36

        I got it. But you will not be, in PREVAIL, for example, will you be looking at lathosterol, desmosterol levels?

    21. JKJohn Kastelein1:51:43

        We have a lot of spare tubes.

    22. PAPeter Attia1:51:45

        Mm-hmm.

    23. JKJohn Kastelein1:51:46

        So that is definitely something that, you know, we have a quite a wish list of things where we wanna look at. Yeah.

18. 1:51:53 – 2:04:17

    APOE4 and cardiovascular disease: remnants, LDL receptor downregulation, and residual risk beyond ApoB

    1. PAPeter Attia1:51:53

       Uh, well, inquiring knuckleheads like me wanna know. Let's close out our discussion, John, with an explanation of the role of ApoE in cardiovascular disease, 'cause it... You know, ApoE gets a lot of attention for what it's doing in the brain. We just had a pretty brief but insightful discussion on that. But I think people are less clear on the relationship of ApoE in the heart. So what, what can we say about atherosclerosis and ApoE?

    2. JKJohn Kastelein1:52:21

       Again, ApoE4 might have been wonderful in during the Ice Age, but now it's bad because it's associated with higher LDL, it's associated with a more pro-inflammatory state, and it's associated with more heart disease. Now, if people have a hard time believing how it's possible that when you're an E4 that you have higher LDL, the explanation is, is that ApoE, of course, sits on VLDL and on VLDL remnants. And ApoE4 actually is a better ligand for the [chuckles] LDL receptor. It's an amazing story, the ApoE story. Better ligand than E32. And so especially the chylomicron remnants and the VLDL, so IDL races into your liver, and that will down-regulate the LDL receptor, and therefore LDL goes up.

    3. PAPeter Attia1:53:13

       Sorry, just to make sure folks understand that, 'cause that's... It's a bit counterintuitive, right? This is a little bit of a paradox.

    4. JKJohn Kastelein1:53:18

       It is totally counterintuitive.

    5. PAPeter Attia1:53:20

       If you have ApoE4 and it's a higher affinity ligand for the LDL receptor, it should mean that ApoE4-generated proteins lead to more rapid clearance of LDL. But if I understand you correctly-

    6. JKJohn Kastelein1:53:38

       No, not LDL. There's no ApoE on LDL.

    7. PAPeter Attia1:53:42

       Y- yes, yeah, okay. So sorry, sorry, yeah. So but-

    8. JKJohn Kastelein1:53:44

       It sits all on the remnants and on VLDL.

    9. PAPeter Attia1:53:47

       I see, and that's the point. So because it's the remnants and the VLDL, those readily get attracted, down-regulate the LDL receptor, so there's less LDL receptor to get rid of LDL. Is that the chain of events?

    10. JKJohn Kastelein1:54:01

        That's the chain of events. But there's much more to E4 than just a little lipoprotein metabolism. There's, like in the brain, there is enough association with other things that are... that you don't want. I mean, this chronic pro-inflammatory state that people like Paul Ridker, you know, has made his life's work of, is also associated with an, with an E4 carriership.

    11. PAPeter Attia1:54:25

        And does the association with ApoE4, in terms of risk, vanish once you normalize for ApoB? Or is there still residual risk based on these other factors, such as inflammation, that exist and persist once you've normalized for Apo3 versus Apo4 in the context of the same LDLC/ApoB?

    12. JKJohn Kastelein1:54:49

        Uh, if you ask Alan Sniderman, he'll say that if you control for ApoB, everything falls away, and that it's the number of p-

    13. PAPeter Attia1:54:58

        Yes, yeah, yeah.

    14. JKJohn Kastelein1:54:58

        You can also look at NMR LDL particles, but that ApoB is really, uh, is really what drives all our statistics in cardiovascular disease. And it's very interesting. For example, you've seen the bempedoic acid data. So the CLEAR Outcomes trial was on the line of absolute LDL lowering versus MACE reduction. So where is the room or the space for the CRP lowering effect if you're on the line? And I really admire Alan. I mean, he has been stoic in his emphasis on ApoB for so long, and finally he gets this kind of vindication now where, where most people would agree that non-HDL and ApoB are a better prognostic marker and a better measure of therapy than LDL cholesterol by the Friedewald formula. But if you ask me personally, Peter, I think that ApoE4 has a few properties that you cannot- completely knock out statistically or biologically with ApoB lowering

    15. PAPeter Attia1:56:03

        Yeah, I don't think that's a big stretch. And, and it might be that Alan's view and that view are not completely at odds, right? It could be that those other effects are small enough that on a clinical level you would need really large sample sizes and lifetime exposure metrics to see the difference. It also could be that there are other amplifying features. In other words, if you take two healthy people, one E4, one E3, with the same ApoB, the risk is relatively similar. But if you take two unhealthy people, type 2 diabetes, NAFLD, profound insulin resistance, one E4, one E3 with the same ApoB, it could be that those other factors create more of a gap between those people on a Kaplan-Meier curve.

    16. JKJohn Kastelein1:56:54

        I think that that's very reasonable, and I, I tend in that same direction.

    17. PAPeter Attia1:56:58

        I mean, I think for me, part of the takeaway here is just that I'm trying hard not to be excited, right? I think that, as I said before, when you're once bitten, twice shy, and I've been a very vocal critic of these trials, meaning the, the, you know, the last 15 years of CETP inhibitors. And I would say that perhaps I've been too harsh, and I've been mostly a critic of the HDL hypothesis, and I've relied on the CETP story along with the Mendelian randomization as my rationale for that. And in reality, I think that that has created a little bit of a blind spot in my eye towards a more pure biologic understanding. And of course, I'm the last person who should have this blind spot because part of my interest is in longevity, and it's clear to me that CETP is a longevity gene, meaning the hypofunctioning variant is as much a longevity gene, if not more, than the hypofunctioning PCSK9. So it's, it's with a tincture of embarrassment that this is sort of my mea culpa to say I have been too down on CETP inhibitors, and I'm very hopeful that obicetrapib not only redeems the field, but also gives me something to be excited about clinically in my practice.

    18. JKJohn Kastelein1:58:15

        Thank you very much for that. And to be quite frank, I was much like you for a long time, and it is... You know, you have to imagine the, the kind of turnaround you have to make in your head. I was a strong believer in the HDL hypothesis until the Roche trial. Then I had to turn around away from HDL back to LDL and ApoB and then understand the blood pressure effect, understand the mistakes in the trials, try to find a drug that didn't have all that baggage. My first New England paper was in 1987. What we did is we measured CETP activity in a coronary angiography trial. In those days, we only had coronary angiography with pravastatin, and it was if you had high CETP activity, you had the worst progression of coronary disease over that two-year trial. So that taught me that CETP was bad, and since that day, I'd wanted to find an inhibitor that was something that I could work with.

    19. PAPeter Attia1:59:15

        It's such a beautiful story. You've articulated it very well, John, and I do hope that at some point you find the time to write kind of a clinical philosophical paper about science, which is we lost our way. We as the field, meaning like, meaning I'm just observing the field, but the field lost its way in this drug based on the wrong biomarker. We were using HDL-C as a biomarker because we didn't have a better biomarker for CETP inhibition, and so the initial insight in the biology was right, but by having the wrong biomarker and failing to understand the mechanism of that, 15 years and I don't know how many billions of dollars, but it's-

    20. JKJohn Kastelein2:00:05

        Oh my Lord

    21. PAPeter Attia2:00:05

        ... probably approaching 10 billion if not more dollars were wasted. And tragically, by the way, it should be noted, some lives were lost, right? We don't want to lose sight of the fact that this resulted in-

    22. JKJohn Kastelein2:00:17

        Absolutely

    23. PAPeter Attia2:00:17

        ... lives lost during these clinical-

    24. JKJohn Kastelein2:00:19

        Yeah

    25. PAPeter Attia2:00:19

        ... trials, and that's the price we do pay as a society to advance knowledge. I don't want to suggest that you can never have a loss of life in a clinical trial, but it's just a sobering story.

    26. JKJohn Kastelein2:00:30

        But if you would forget about everything that happened before, what we want nowadays is strong Mendelian randomization evidence. We want lots of phase 1 and phase 2 trials to show that there are no some bizarre side effects, and then you gently go into a phase 3 with a DSMB and everything, and you don't do a phase 2 where you see a blood pressure effect and you say, "Oh, what the hell?" And you know, well, that would not be possible today anymore, fortunately.

    27. PAPeter Attia2:01:01

        And that's a good thing. Yeah.

    28. JKJohn Kastelein2:01:02

        Yes.

    29. PAPeter Attia2:01:03

        Well, John, I'm really glad that Tom reached out to you to ask if you'd be willing to be on this podcast. I'm grateful that you were able to accommodate my time today, and thank you again for the work you're doing, and I'm very excited to follow this story, as will the other listeners today.

    30. JKJohn Kastelein2:01:18

        We'll have data on Alzheimer in the summer. Really thrilled to understand what's happening in the brain.

Episode duration: 2:04:17