The Peter Attia Drive - #10 - Matt Kaeberlein, Ph.D.： rapamycin and dogs — man's best friends？ — living longer, healthier lives and turning back the clock on aging, and age-related diseases

00:00:00.000 Hey everyone, welcome to the Peter Atiyah Drive. I'm your host, Peter Atiyah.

00:00:10.160 The Drive is a result of my hunger for optimizing performance, health, longevity, critical thinking,

00:00:15.600 along with a few other obsessions along the way. I've spent the last several years working with

00:00:19.840 some of the most successful, top-performing individuals in the world, and this podcast

00:00:23.620 is my attempt to synthesize what I've learned along the way to help you live a higher quality,

00:00:28.360 more fulfilling life. If you enjoy this podcast, you can find more information on today's episode

00:00:33.000 and other topics at peteratiyahmd.com.

00:00:41.320 In this podcast, I'll be speaking with Matt Kaberlin at the University of Washington. I met Matt about a

00:00:47.200 year and a half, maybe two years ago, through David Sabatini, who some of you may be familiar with just

00:00:52.480 based on all of his remarkable work around mTOR and rapamycin. Matt's recognized globally for his

00:00:58.340 research in the biology of aging. He got his PhD from MIT in the lab of Lenny Garante, who's produced

00:01:04.920 a number of notable folks in this field. He went on to do his postdoc at the University of Washington,

00:01:10.100 and after completing his postdoc, he has remained there and continues to run a fantastic lab.

00:01:16.340 In this episode, we're going to talk about his experience as the director of the Dog Aging Project,

00:01:20.700 which, as its name suggests, focuses on the animal model of dogs for its research. Now, of course,

00:01:26.120 this is really interesting because while fruit flies and yeast and mice are interesting, dogs are

00:01:31.940 obviously much closer to us. Of course, these dogs, because they're pets generally, have something

00:01:37.420 really unique to us that virtually no other research animal has, which is they share our environment.

00:01:43.900 This puts them in a pretty unique spot that even the rhesus monkeys studied in the NIA Wisconsin

00:01:49.340 project didn't have going for them. Now, if you haven't already done so, I'd recommend listening

00:01:54.880 to the podcast that I did with David Sabatini because that will get you some of the background

00:01:59.800 on mTOR and rapamycin. I think we get a little technical in this podcast, but I suspect it's

00:02:05.940 something that if you've listened to that other one with Sabatini, you'll have the background to

00:02:09.440 follow. We talk about a bunch of other things that people seem to be interested in as well,

00:02:12.580 beyond rapamycin and aging. We also talk about NAD, probably get into sirtuins a little bit.

00:02:17.560 The other thing I really enjoy about speaking with Matt, and I had so many discussions with Matt,

00:02:21.880 you know, over meals and stuff where I think to myself after, man, I wish that was recorded

00:02:25.620 because one, I want to be able to hear it again. And two, it's the kind of stuff that people are

00:02:29.480 always asking me. And I think Matt just has such an amazing way of thinking about this stuff.

00:02:34.480 Matt's work is really remarkable because it's actually focusing on health span. It's easier in

00:02:38.440 some ways to study health span because you can study it over a shorter period of time. And in particular,

00:02:43.240 when you look at the cardiomyopathy model, meaning it's a type of heart failure that

00:02:47.560 dog's experience, and you look at how rapamycin can improve ejection fraction,

00:02:51.560 you can get these answers in a really quick period of time. In fact, much quicker than I think Matt

00:02:55.660 even expected. And we talk of course about cancer, heart disease, and Alzheimer's disease. So

00:02:59.680 if you're interested in rapamycin, if you're interested in mTOR, if you're interested in longevity,

00:03:03.920 I think you're going to find this interview very interesting. It was actually recorded initially in

00:03:08.620 December of 2017 as part of the interview series I was doing for my book. And we may at some

00:03:13.600 point throw up the video as well. You can find a lot more information in the show notes,

00:03:18.720 of course. And without further delay, here's my conversation with the remarkable Matt Kaperler.

00:03:27.120 Matt, thank you so much for being here. It's really fantastic to be able to talk with you in this

00:03:32.380 format. I've been a huge fan of your work for many years now. Obviously, I know quite a bit about your

00:03:40.660 background, but I think it'd be great for the listeners to kind of get a sense of, you know,

00:03:44.140 what did you study in college? How'd you end up doing your PhD? Where'd you do it? And most of

00:03:48.180 all, why'd it get you where you are? Sure. So I got undergraduate degrees in biochemistry and

00:03:53.620 mathematics at Western Washington University. And then I went to graduate school at MIT,

00:03:59.340 the biology department there. And my background up till that point had been in biophysical chemistry.

00:04:05.420 So I really went to graduate school thinking I was going to work on structural biology or

00:04:09.900 x-ray crystallography or something like that. And I heard a seminar by Lenny Guarenti,

00:04:15.080 who I ended up doing my PhD thesis with during my first semester at MIT, where he started talking

00:04:20.540 about how his lab had begun working on the genetics of aging and trying to understand,

00:04:27.060 you know, what are the factors that influence the rate of aging? And this was in yeast. That's what

00:04:31.720 his lab worked on at that point. And it was really almost like an aha moment where it just clicked

00:04:38.220 with me that it was really fascinating that you could use genetics and molecular biology and

00:04:44.340 biochemistry to study something as complicated and fundamental as the biology of aging. And so I got

00:04:50.660 fascinated by the topic, ended up going and talking to Lenny, and then subsequently doing my thesis

00:04:57.800 research with him. So it was really that moment of hearing him talk my first year in graduate school

00:05:03.040 that got me interested in aging. And that's what I've been fascinated by and passionate about since

00:05:07.900 then. And so then I went on and did a postdoc with Stan Fields at the University of Washington,

00:05:12.860 also working on aging, and then ultimately started my own lab. So it's been almost two decades now that

00:05:18.580 I've been working on this problem. And if my memory serves me correctly, when you were in Lenny's lab,

00:05:24.120 who's obviously no stranger to probably people who are listening to this, you worked on sirtuins,

00:05:28.700 correct? Yeah. So my thesis work was actually the work that first showed that if you activated or

00:05:36.220 overexpressed a sirtuin, in this case, it was sirtu, the founding member of the sirtuins in yeast,

00:05:42.440 that you could extend lifespan and slow aging. And obviously, we were very excited about that,

00:05:47.500 especially when a postdoc who was in the lab at that time, Heidi Tissenbaum, about a year later,

00:05:53.440 showed that you could also overexpress the worm version of sirtu and extend lifespan in C. elegans.

00:06:00.060 And so that was really the first example of a conserved genetic modifier of lifespan

00:06:06.340 across two widely evolutionarily divergent organisms. Because it was a relatively similar

00:06:12.200 version of sirtu. Yeah. So more different than the DAF stuff that we saw in C. elegans versus

00:06:17.140 the IGF that we would see in some of the higher mammals. Right. So at that time, we knew that DAF

00:06:23.960 2 could extend lifespan in C. elegans. I don't think we knew it was working on a similar mechanism.

00:06:28.940 That's right. Yeah. Now, I think one of the things that makes your work so interesting is that you work

00:06:34.380 on dogs. Right. How did you end up doing that? Yeah. So that's relatively new in my career. It was

00:06:41.020 about four years ago when Daniel Promisloe moved from the University of Georgia to the University of

00:06:46.560 Washington. And Daniel had started thinking about dogs as a model to understand how genetic and

00:06:55.200 environmental factors influence the aging process. And I'm a dog person. I've always had dogs. I have

00:07:01.300 three dogs now. That qualifies. Yeah. Right. And I had never really made the connection between

00:07:07.020 dogs as companion animals and dogs as a model for understanding aging until I talked with Daniel

00:07:14.200 about this. And through those conversations, it occurred to me that not only could we understand

00:07:19.760 genetic and environmental modifiers of aging, but our pets could actually serve as a transformative next

00:07:27.640 step in testing some of these things that we know work in laboratory mammals like mice, but have never

00:07:34.760 yet been taken outside of the laboratory. And so it was really that sort of connection in my mind about

00:07:39.740 three years ago that propelled me to really move forward with the dog aging project. One major goal

00:07:46.120 of which is to actually do this, to take some of the interventions that we know extend lifespan and slow

00:07:52.780 aging in the laboratory and test whether they have that effect in the real world in a larger mammal

00:07:59.640 that's also very socially relevant. I think that that's an important aspect of this. People love their

00:08:04.700 dogs. If we can actually slow aging in people's pets, that's going to have a huge impact both on

00:08:11.080 the quality of life for the pets and the owners, but also the way that people think about the biology

00:08:16.560 of aging. They're going to believe it a lot more if they see that their dog is living longer and aging

00:08:22.300 more slowly than just reading an article in the New York Times or wherever. Right. Right. And it's,

00:08:27.820 you know, I sort of, when I talk about this, I generally talk about these four classes of eukaryotes going

00:08:33.540 from, you know, yeast to flies, worms, and mammals. But usually when we talk about mammals, we're

00:08:38.500 talking about mice. Absolutely. And they have a couple of problems at least, right? One problem is

00:08:44.140 they're generally, especially if they're inbred, I mean, they're homozygous at all loci. So you have

00:08:48.700 to question their applicability to us in terms of their susceptibility to disease. But then the second

00:08:54.060 point, which you alluded to, which I think the dogs might be the first quote unquote laboratory study

00:09:01.140 that gets out of this is they don't live in our environment, right? The mice, that is. Right.

00:09:05.100 Right. So the dog truly lives in the environment you and I live in, which at least in theory should

00:09:12.060 be the environment we care most about preventing aging in. That's right. I think that environmental

00:09:16.080 variation is hugely important. The genetics is also important. And dogs are, so there's a couple

00:09:22.560 things that are worth mentioning. Dogs have some of that same genetic homogeneity if you look within

00:09:30.000 individual purebred breeds. But we also have many purebred breeds that are widely divergent,

00:09:38.100 both genetically and morphologically. All you have to do is look at a Chihuahua on a Great Dane,

00:09:42.400 right? To see that divergence. But we also have this really interesting and complex mixed breed

00:09:48.620 population. That's a combination of all these different genetic variants that each breed has.

00:09:53.340 So we have kind of the best of both worlds in the sense that if you want to do a study in a

00:09:58.520 relatively genetically restricted background, you can do it in a specific breed or set of breeds.

00:10:04.220 But if you want to capture that genetic variation, all you have to do is look in the mixed breed dogs.

00:10:08.600 But the environmental part of this, I think, is probably the most important from a,

00:10:13.600 what are we going to learn from dogs that we can't learn from mice perspective? As you said,

00:10:18.400 dogs really share our environment to a greater extent than any other animal,

00:10:22.040 maybe with the exception of cats. And so they're drinking the same water. In fact,

00:10:27.280 may not even be drinking the same water. Most people-

00:10:29.580 Maybe drinking worse water.

00:10:30.580 Yeah, right. Are not going to give their dog bottled water, right? But they're breathing the

00:10:33.840 same air. If you smoke, they're experiencing that, or someone in the house smokes, they're experiencing

00:10:38.620 that secondhand smoke. So they really do capture most of our environment. The diet is about the only

00:10:43.920 place where dogs don't quite have the same sort of nutritional diversity that people do. But with

00:10:50.500 the exception of diet, the environment is pretty close.

00:10:53.260 Now, we have a pretty good idea of how we die, right? I could actuarially map out how you and

00:10:58.940 I are going to die based on our age and a whole bunch of other factors. Again, it's probably an

00:11:03.660 oversimplification because of the species diversification. But as a general rule, how do dogs die?

00:11:08.900 Yeah. So I'll make one quick comment on that. First of all, that I think you're right that we

00:11:14.160 do know what diseases most often kill people. And that's important information. I'm not sure it's

00:11:21.200 the most important information because what people die from does not always equate to what they die

00:11:26.600 with. Especially today, most people are dying with- With chronic diseases in place. Yeah, multiple

00:11:31.860 comorbidities, right? Yeah. So just because you may die from heart disease, it doesn't mean that you

00:11:37.140 didn't have kidney disease or something moving towards kidney disease or diabetes. So I think

00:11:42.220 it's important to appreciate that you can have multiple diseases and only one of them is probably

00:11:48.240 going to kill you. Having said that, it is useful to think about, do dogs die from and with the same

00:11:56.820 age-related diseases that people do? And the answer is, in general, the equivalency is pretty good.

00:12:02.100 So dogs get all of the same age-related diseases that people do. They don't get them at the same

00:12:07.380 frequency necessarily. So one big difference is in dogs, there's actually relatively little

00:12:14.860 vascular disease, which is a major killer in people. So specifically atherosclerosis,

00:12:20.660 whether it be peripherally or cardiovascularly, doesn't seem to be as prevalent. Does not seem to be

00:12:25.640 as prevalent. But there are breeds that die from heart disease and certain forms of heart disease.

00:12:32.100 Cancer is probably the most common cause of death in dogs. And big dogs tend to get more cancers than

00:12:39.040 small dogs. But across all dogs, cancer is probably the number one cause of death. Kidney disease is a

00:12:44.180 major cause of death in dogs. Is the etiology of that kidney disease related to blood pressure? Is it

00:12:49.900 related to some other nephrotic syndrome that's otherwise unidentified? I don't know the answer to that.

00:12:55.640 Yeah. I don't, I don't have the veterinary background to answer that. One of the things

00:12:59.480 that's interesting in dogs is that there's, it seems to be a little bit of a debate whether dogs

00:13:03.440 really get Alzheimer's disease. They clearly get dementia. Some dogs do. And they clearly show

00:13:09.640 cognitive decline. It's not completely clear whether they get what would be clinically diagnosed as

00:13:14.960 Alzheimer's disease in people. Apparently they do accumulate a beta in the brain, whether they get

00:13:20.140 the plaques and tangles still seems to be a little bit up in the air. And there are people studying

00:13:23.700 that, but at least at the level of cognitive dysfunction and dementia, it's clear that dogs

00:13:29.500 experience that with age as well. So for the most part, they do develop the same age related diseases.

00:13:34.700 The increase in risk for those diseases goes up essentially exponentially, just like in people,

00:13:39.920 but the relative prevalence of specific diseases is not always the same and can also be breed dependent.

00:13:46.420 So there, it is certainly the case that within purebred dogs, different breeds have different

00:13:52.620 predispositions to certain diseases, which is exactly what you'd expect based on the genetics.

00:13:57.720 So basically they're probably getting a little more cancer, significantly more renal failure as a

00:14:02.820 proximate cause of death, as opposed to dying with renal insufficiency, which to your point,

00:14:06.220 I think a lot of humans do. They're getting a lot of heart disease, but it sounds like it's more

00:14:09.700 cardiomyopathy and or valvular disease, but not atherosclerotic or ischemic disease.

00:14:15.000 And they unfortunately die of accidents just as humans do, which would probably be in the top 0.92

00:14:19.080 four for humans and dogs, I'm guessing. Yeah. And actually that's an interesting point. You're

00:14:23.920 right. What I haven't seen great data on is the age distribution of death due to trauma.

00:14:30.200 I suspect it's going to be mostly younger dogs, but I don't know for sure. So that's actually an

00:14:35.680 interesting question. It is absolutely true. The other thing that is worth noting about dogs is it's

00:14:40.660 actually the case that most pet dogs don't die from an age-related disease. They die from euthanasia,

00:14:47.380 which is a difference between dogs and people. Yes. That's a great point.

00:14:51.460 Well, let's turn the discussion now to what you do about this. I've been a big fan of rapamycin and

00:14:57.620 TOR and that entire pathway for several years now. It's become almost an obsession. One can have

00:15:04.440 an obsession that's not pathologic. And I think what attracted me to your work a couple of years

00:15:10.680 ago, probably David Sabatini pointed me in your direction. So maybe three years ago, it was actually

00:15:16.160 after the manic paper came out in, which was around 2014, right? Which was sort of the first really

00:15:22.560 interesting look at the human data. I want to come back to that paper as it relates to immune function

00:15:28.060 and stuff. But maybe give us a sense of how you decided that the next logical step was to actually

00:15:35.340 test the God molecule, as I like to call it, the rapamycin in this context. In dogs, yeah. Yeah. So

00:15:42.660 again, this was all happening about three years ago when I first sort of made the leap to thinking that

00:15:50.180 we could test interventions in dogs. And actually that's not, at least for me, was not an immediate,

00:15:55.960 it wasn't immediately obvious to me that we should. So you really have to think, as soon as you start

00:16:00.980 talking about bringing trial of a drug out of the lab and into the real world, whether it's the human

00:16:07.640 clinic or the veterinary clinic, in the context of aging, you really have to start to think about

00:16:13.020 safety and side effects, right? Because there is a very low tolerance for side effects when you're

00:16:19.440 talking about treating a healthy person or dog. So that was the first thing that I had to really

00:16:24.860 come to grips with is, could we do this safely? Because to be clear, you were thinking,

00:16:29.800 we're not going to take dogs that are already sick. That's right. We were going to take healthy

00:16:33.320 dogs. That's right. I mean, for me, as somebody who is fundamentally interested in the biology of

00:16:38.140 aging, the intent is to slow aging in people before they get sick, right? To keep them healthy

00:16:45.560 longer. So in many ways, it's the opposite of traditional medical approach, biomedical research,

00:16:51.580 right? Where normally, historically, we wait until people are sick and then we try to cure their

00:16:57.460 disease. This is the reverse of that, right? So that's right. So because we are talking about

00:17:03.060 intervening in a healthy person or a dog, the tolerance for side effects from a regulatory

00:17:09.540 perspective or even just public perception is very low when you're talking about a healthy person.

00:17:15.520 Now, you know, the way I view that, first of all, I think that is we need as a society and within the

00:17:21.660 scientific community to have a discussion about this, because I think that we need to recognize

00:17:27.860 that a healthy 70-year-old is not the same as a healthy 30-year-old, right? And we know what's

00:17:33.720 going to happen if we don't do anything about aging. So my personal view is that there should be a

00:17:38.000 tolerance for some level of risk if the outcome is going to be 10%, 20%, 30% more time spent in good

00:17:47.360 health. That's a discussion that we haven't had either at the regulatory level or at the

00:17:52.320 society-wide level, but I think we need to have. So I digress a little bit, but I had to go through

00:17:57.980 this. I think that's one of the most important points you could make. And I sort of think a lot

00:18:02.120 about this because my colleague and I were discussing this the other day, which was, you know, you take this

00:18:06.540 oath at the end of medical school, the Hippocratic Oath, the first thing you learn to say is, first,

00:18:10.100 do no harm. And I think the spirit of that is excellent, but I also think it's highly impractical

00:18:15.200 in a world where it forces you into binary thinking, which is we will only undertake

00:18:21.020 interventions that are guaranteed to have no harm, and otherwise we will do nothing regardless of the

00:18:26.920 outcome, which, of course, is not practical. We live in a probabilistic world where the probability

00:18:31.320 of harm is not zero or one, but rather it's a continuum between zero and one. And you really need to

00:18:36.040 take a risk-adjusted approach to outcomes. So I mean, I think that's an excellent point.

00:18:41.260 Yeah, that's exactly right. So I had to kind of go through that thought process in my own head,

00:18:44.440 though, and especially thinking about moving into pet dogs, where because of the way that

00:18:50.840 many people feel about their pet dogs, you have to be as sure as you can that you're not going to

00:18:57.280 hurt anybody's pet in a trial like this. They like them more than their friends.

00:18:59.640 Yeah, absolutely. Yeah, people love their dogs, right? A lot of people feel similarly about their

00:19:06.540 dogs as they do about their kids, right? So you kind of think that's kind of the way I thought

00:19:09.640 about it is, could we do this in somebody's child? And so with rapamycin, there is a perception out

00:19:16.120 there that I don't share, but there's a perception out there that rapamycin has lots of side effects

00:19:20.880 based mostly on the human clinical literature. Which, to be clear, is generally in transplant patients.

00:19:26.260 Transplant patients taking high doses and taking lots of other drugs. Yeah. So sick people 1.00

00:19:30.780 taking a high dose of rapamycin in combination with other medications, right? And it does have

00:19:35.860 side effects. There's no question about that. But one of the things that's come out of my own

00:19:40.620 research and other research in the field is that both the benefits and the side effects are strongly

00:19:46.660 linked to dose. So one question was, is there a dose of rapamycin that will have beneficial effects

00:19:55.200 in the context of healthy aging without significant side effects? That was an unknown. I think that

00:20:01.160 we're getting to the point where we're pretty sure that that's the case. At least you can get some of

00:20:05.660 the benefits of rapamycin and we can talk more about the data that support that. But at that point,

00:20:10.260 it really wasn't clear. And when you say the dose, do you think that the peak or the trough

00:20:16.120 play a bigger role in toxicity? Yeah. So that's still an unknown, but I think the data that are out

00:20:22.560 there suggest that the trough levels are most strongly correlated with side effects. Now,

00:20:29.880 what is most strongly related to healthspan or lifespan? There's no data, as far as I know.

00:20:35.680 And this is an area where I think there's a lot of work that could be done and should be done

00:20:40.520 exploring more broadly, even in laboratory animals and mice, the sort of dose response and dose timing

00:20:47.520 space for where do you get the biggest effects on lifespan or specific measures of health?

00:20:55.000 Where don't you get any effects? Can you uncouple, say, the improvements in heart function from the

00:21:00.400 improvements in immune function by changing the dose or the timing? There really has been very

00:21:05.580 little done on that. Now, maybe for the listener who's not familiar with the pharmacokinetic

00:21:08.840 discussion, let's explain maybe trough and peak. How do we think about those doses?

00:21:13.180 Yeah. So when you give a medication, say a rapamycin pill, when a person takes a rapamycin pill,

00:21:18.760 there will be a rapid increase in the levels of the drug in the blood. And if they don't take

00:21:24.020 another dose, then the drug will start to get cleared and it will go down. And so if you're

00:21:28.420 taking a pill every day, as soon as you take the pill, the blood levels go up and then it starts to

00:21:32.960 get cleared. And then the next day you take the pill, it goes up and then it starts to get cleared.

00:21:36.740 So if you were to take the pill every other day, it would go down further before you get that spike

00:21:42.800 again. So the spike, the top of the spike is the peak level, the bottom before you take the next

00:21:47.640 dose is the trough. And so the little bit of data that's available, and again, there's not much data

00:21:53.820 for different doses of rapamycin in people who are not also taking other drugs. So the combination

00:22:01.220 here is both dose and rapamycin as a monotherapy. And so really the only study that I know of that

00:22:07.520 looked at this really at all actually didn't even use rapamycin. They used a derivative of

00:22:12.440 rapamycin called RAD001 or Everolimus. That's the manic study that we talked about. So there's

00:22:17.720 really no good data in people on different doses of rapamycin as a monotherapy in healthy people.

00:22:24.800 So we're kind of stuck looking at the data that we have. So in Joan's study with Everolimus or RAD001,

00:22:32.600 they gave the medication to healthy elderly people and they tested three dose and delivery

00:22:39.460 combinations. So one of them was, I believe, 20 milligrams once a week. One of them was five

00:22:46.180 milligrams once a week. And one of them was one milligram a day. Right. And this was looking in

00:22:51.800 the context of immune function as measured by a flu vaccine response. So the outcomes were that I think

00:22:59.260 at all three doses, they saw evidence for improved vaccine response, which was consistent with prior

00:23:05.840 data in mice, that immune function is improved by rapamycin. Pause for a moment. That's still a bit

00:23:10.860 counterintuitive to the layperson. When I say the layperson, I mean like the layperson who still

00:23:16.440 thinks about rapamycin. Or actually lots of physicians. Sure. Because we think of this as an

00:23:20.560 immune suppressant. That's right. Yeah. And yet they took this drug in monotherapy under a different

00:23:26.080 dosing schedule than a transplant patient would take it. And we saw an improvement in their T cell

00:23:31.060 function. The same cells that we tend to knock out in a transplant patient. Yeah. So let's come

00:23:37.220 back to that. Because that I think is an important issue, but it's complicated. Just to come back to

00:23:41.760 the trough levels and side effects. So they saw evidence for efficacy with every delivery method.

00:23:48.620 Although they got, I think, the best efficacy at the five migs once a week. But they also had the 1.00

00:23:54.040 lowest side effects. And the most side effects were the ones every day or the twenties once a week.

00:23:59.140 I think it was at the twenties once a week, but I can't remember for sure. It was pretty comparable.

00:24:03.400 So the first thing to say is none of the side effects were bad by clinical standards. None of

00:24:08.340 the people dropped out of the study because they were taking the drug, which is, I think, a pretty

00:24:12.340 good indication for how tolerable the drug is. So even though they did detect some side effects,

00:24:18.500 they really were not serious. They weren't even serious enough that they were uncomfortable and

00:24:23.620 people stopped taking the drug. So I think that that's really the only evidence that we have,

00:24:28.180 that I know of, that it's really the trough levels that drive side effects. So I kind of think that's

00:24:34.900 probably true, but I'm not confident. And I really think we need more data to know for sure.

00:24:40.420 It's certainly true in other classes of drugs. I mean, when you look at gentamicin, for example,

00:24:44.820 and the ototoxicity or nephrotoxicity, it's a trough problem, not a peak problem.

00:24:49.380 That's why you have to make sure the patients clear it before you redose it.

00:24:52.980 Right. So there are other reasons to believe that as well. Yeah. So now coming back to this immune

00:24:57.260 function question. There was a washout.

00:24:59.760 There was. So there's a couple of things, again, there to consider. So the data suggesting that

00:25:04.500 rapamycin can act as an immunosuppressant, again, is almost exclusively based on very, well, I shouldn't

00:25:13.180 say very high doses, higher doses than were tested in the Novartis study. In people who are also taking

00:25:20.260 other drugs that probably are true immunosuppressants.

00:25:23.540 Typically at least two, if not three other drugs in that cocktail.

00:25:26.520 So it's really not clear to what extent rapamycin as a monotherapy in healthy people

00:25:33.180 has immunosuppressive properties. And the data in mice, I would say is mixed. It really seems to be

00:25:39.320 the case that for some forms of immune challenge at high doses, rapamycin can enhance susceptibility

00:25:48.120 to infection or other forms of immune challenge. It enhances function. But again, those studies are

00:25:53.380 almost always done at very high doses of the drug that are even much higher than you would give to

00:25:58.460 a person. So it's an unknown. Now, what seems to be the case, both in mice and people, is that

00:26:05.000 short-term treatment with rapamycin in an old mouse or an old person, followed by a two-week washout

00:26:12.580 where they stopped taking the drug. When you then test immune function, at least as measured by a vaccine

00:26:18.560 response, you get a better response. So one model would be that the treatment with rapamycin is restoring

00:26:28.240 immune function in an aged animal, probably through enhanced stem cell function, although I don't, I think

00:26:36.520 that that hasn't really been demonstrated clearly. And you might need that washout period. If there is

00:26:42.960 an immunosuppressive effect, you might need that washout period to be able to see that rejuvenation

00:26:47.680 and immune function. Again, that's really speculative though, because nobody's done.

00:26:51.720 In Joan's paper, did they do, I'm sure they didn't actually, now that I think about it,

00:26:56.960 they didn't have enough people in the study. It would have been very interesting to have seen the

00:26:59.720 immune challenge without the washout and the subset.

00:27:01.980 Now, I was just going to say, nobody's done it in either mice or in people. That actually would be

00:27:06.800 a fairly easy experiment to do in mice. The problem is you would never get an NIH study section to fund

00:27:13.900 that experiment because we already kind of know the answer that rapamycin works. They wouldn't be

00:27:18.720 viewed as an, even though it's really important from a translational perspective, it wouldn't be

00:27:23.780 viewed as innovative or important enough for somebody to fund a grant to do it. So it's an unknown,

00:27:29.680 and I don't know how long it'll be till we actually get the answer to that question.

00:27:33.460 Now, bringing it back to the dogs, one of the challenges, of course, in leaping from mice to

00:27:39.560 dogs is mortality becomes difficult to study. You have an animal that lives a lot longer,

00:27:46.660 whereas mice, you know, you can get a longevity answer in, you know, months if you select them

00:27:51.360 correctly. And dogs, if you want to study them for true hard outcome of, you know, death,

00:27:56.680 so you pick a proxy. Well, you can.

00:28:00.440 Yeah. I'm saying, I'm saying like the shortest path would be, let's look at organ function or

00:28:05.720 something else.

00:28:06.600 Right. So what we did, so we've done one study where it was a 10 week study of rapamycin in

00:28:13.120 middle-aged, healthy companion dogs. And we chose heart function as our short-term measure. And that

00:28:19.920 again was based on mouse data where two different, actually three different labs now have shown that if

00:28:25.360 you take a 20 to 24 month old mouse, that's maybe the, that's like a 40 year old. No, it's more like

00:28:30.880 a 60 to 65 year old person that if you just look at the heart of a 24 month old mouse compared to,

00:28:36.880 let's say a six month old mouse, you can see declines in heart function, just like you can in

00:28:42.660 people. And, and the parameters that have been studied with respect to rapamycin specifically

00:28:47.760 are mostly measures of left ventricular function. So ejection fraction, fractional shortening,

00:28:53.080 things like that. And this is done by echocardiography. So it's relatively non-invasive.

00:28:57.120 So you can see a decline in function with age. And what, what has been seen now in three independent

00:29:02.460 studies is that six to 10 weeks of treatment with rapamycin is enough to cause those measures of

00:29:09.780 heart function in the old mouse to go back about halfway to what you would see in a young mouse.

00:29:16.180 So it doesn't bring you all the way back to a teenager, but it gets you back to, you know,

00:29:20.540 maybe a 35 year old heart. If you're doing this sort of mouse to human equivalency.

00:29:24.580 And just to be clear, that was how many weeks?

00:29:26.800 So the studies have varied is between six and 10 weeks. All of them saw improvements. It's not clear

00:29:31.240 whether you get bigger improvements by longer treatments. So it's kind of in that range.

00:29:36.800 And did these animals also require a washout to see the benefit?

00:29:39.760 No washout. So these are measures of heart function taken while the mice are still on the

00:29:44.300 rapamycin.

00:29:44.820 And these animals were dosed daily?

00:29:47.060 So yes, all three of those studies used the encapsulated rapamycin in the diet. So they

00:29:54.040 were getting the drug daily and it was not, and this is where it gets a little bit complicated

00:29:57.800 because I'd say 85% of the studies on aging or age-related functional measures in mice with

00:30:04.480 rapamycin have been done with an encapsulated form of rapamycin in the diet. We call it E-RAPA.

00:30:09.840 And so that's different from a pill, right? Because mice are going to eat throughout the

00:30:15.020 night. So they're probably not experiencing exactly the same pharmacokinetics that you

00:30:19.560 would experience from a pill or the other kinds of experiments that people have done in

00:30:23.640 mice have been injections where you get, you know, this rapid peak in the drug and then

00:30:27.540 a pretty steep drop off as the drug starts to get cleared. Nobody's ever done the, like

00:30:31.880 the 24 hour measures of rapamycin blood level on the mice getting E-RAPA.

00:30:37.120 So I don't know how different it is, but it's probably different in the sense that it's

00:30:41.100 probably a more stable level of rapamycin in the blood for a longer period of time.

00:30:46.180 I mean, I would expect that the E-RAPA animals are going to have lower peaks and higher troughs.

00:30:50.680 I mean, that would be...

00:30:51.320 That would be the expectation. Yes. But I don't know that anybody's ever really carefully

00:30:55.140 looked at that. So having said that, I think all three of the studies that looked at heart

00:31:00.760 function in mice used the E-RAPA and it was a pretty low dose of the E-RAPA.

00:31:05.320 How many milligrams per kilogram? It's 14 parts per million in the food. I don't remember off the

00:31:11.680 top of my head what that works out to in milligrams per kilogram. I feel like it's in the, about the

00:31:16.700 two range. Wow. But much. Well, yeah, but again, you have to, you have to keep in mind that you

00:31:23.560 can't translate the dosing across species very well. Two milligrams per kilogram in a mouse is not

00:31:29.300 going to be the same as two milligrams per kilogram in a person or it doesn't have to be.

00:31:32.880 Because the dogs and the humans, I suspect, are probably a lot closer than the...

00:31:36.500 You would think so. Although I don't know that there's actually good evidence to support that.

00:31:40.860 So certainly in terms of body size, if you're talking about a bigger dog, they're closer. In

00:31:44.580 terms of metabolism of the drug though, that I imagine could vary quite a bit from species to

00:31:49.420 species. I don't know. That's a good question. I'm pretty sure it's not renal.

00:31:53.600 From P450. Yeah. It's going to be in the liver. So how did you ultimately come up with both a dose

00:32:02.200 and a schedule of delivery for your first trial? So this again goes back to this question that I

00:32:08.180 was asking myself about, can we do this safely? And there's very little data in dogs on rapamycin at

00:32:15.260 all. And just like in people, there's almost none on rapamycin as a monotherapy in healthy

00:32:21.420 animals. So I started basically digging and talking to as many veterinarians as I could to find out

00:32:29.280 what was known. And I was very fortunate to actually be able to get in touch with a veterinary

00:32:35.300 group at the University of Tennessee who was studying rapamycin in dogs who had had hemangiosarcoma

00:32:42.420 of the spleen. And so they had done the pharmacokinetics and had developed a dosing strategy

00:32:48.360 that they feel is extending life expectancy in dogs who have had hemangiosarcoma of the spleen

00:32:56.480 and where they weren't seeing side effects. And they had had some of their dogs on rapamycin for

00:33:01.140 more than a year. So we were pretty confident, or I was pretty confident after talking to them,

00:33:06.820 that if we took their dosing strategy, that we're unlikely to see any significant side effects over

00:33:12.140 a 10-week period. And so I said that cardiac function was the functional endpoint that we

00:33:16.960 were using. Really, the goal of that 10-week study, though, was to confirm...

00:33:21.080 This is a phase one, basically.

00:33:22.220 Yeah, that we could do this, that we could get people to participate, they would actually give

00:33:25.460 their dog the medication, and that we didn't see any significant side effects. So I was pretty

00:33:29.780 confident that based on their dosing and delivery protocol, we at least wouldn't see any severe

00:33:35.720 side effects in a 10-week period.

00:33:37.120 Did NIH fund that study?

00:33:38.480 Well, not directly. So we got a little bit of NIH money because we have one of these Nathan

00:33:44.260 Schock Centers of Excellence in the Biology of Aging, and we had a small amount. It was

00:33:49.860 on the order of a few tens of thousands of dollars left over from the prior year. And so I asked

00:33:56.220 Felipe Sierra, who's the head of the Division of Aging Biology at NIH, if we could use that surplus

00:34:02.040 specifically for this project. And he was kind enough to agree to that. But we didn't write

00:34:07.460 a grant and get a grant.

00:34:08.260 But you sort of bootstrapped this thing.

00:34:09.660 Yeah. So actually, I paid for this study mostly out of funds that I had gotten because I was

00:34:15.720 recruited to get a job somewhere else. And as part of my retention package, the University

00:34:20.120 of Washington gave me some money that I could spend on whatever I wanted to. And I thought

00:34:24.540 this was a good way to spend the money.

00:34:26.280 So what did that 10-week study show?

00:34:28.120 Right. So first of all, I'll go back to the dosing before I tell you that.

00:34:32.080 Yeah. So I talked to this group at the University of Tennessee. And so their strategy was 0.1

00:34:38.700 milligrams per kilogram given three times a week. Monday, Wednesday, Friday. And that made

00:34:45.500 sense to me in light of what we were just talking about from the manic paper, where it seems like

00:34:52.000 if you give an extra day to let the trough levels get down, it made intuitive sense that

00:34:55.960 that might also...

00:34:56.380 Yeah. So I don't know that that's the case, but it made sense. And so that was the dosing

00:35:01.780 strategy that we went with for our highest dose. And then we also tested a dose that was half of

00:35:06.760 that. So 0.05 milligrams per kilogram, again, Monday, Wednesday, Friday. So we enrolled dogs

00:35:12.880 into this study. This was a very small study at a private veterinary specialty clinic in the Seattle

00:35:19.520 area. And so the dogs had to be at least 40 pounds and at least six years old, and they could not have

00:35:27.180 any preexisting conditions. So again, this is a study of healthy aging. We wanted healthy dogs coming

00:35:33.540 into the study. Six years old, because at that weight range, we figured that that would be roughly

00:35:40.680 the human equivalent of 55 years, maybe. Big dogs age faster than small dogs. So that's why we had the

00:35:47.100 weight criteria. So we wanted to have a population where we expected there would be the potential for

00:35:52.400 some age-related functional decline, but that we wouldn't get a high proportion of dogs...

00:35:59.100 Yeah. You don't want to be too close to the edge of the cliff.

00:36:00.600 That's right. That were really sick. So one thing we found, though, that was unexpected was that about

00:36:06.660 20% of dogs in that age and weight range actually have asymptomatic heart disease that you will see if

00:36:13.320 you give them an echocardiogram, but you won't detect from a stethoscope exam. And that was,

00:36:18.620 in hindsight, it kind of makes sense, right? Again, heart function is going to go down with age,

00:36:23.340 and where you decline, like what is the clinical threshold that we call disease,

00:36:28.900 is sort of a moving target sometimes, right? If a vet with a stethoscope detects a heart murmur,

00:36:34.220 and then they give the dog an echocardiogram, and they see regurgitation, they'll call that heart disease.

00:36:38.520 If they don't hear a heart murmur, nobody's going to give their dog an echocardiogram, right? It's

00:36:43.180 expensive. Yeah. And I'm not a cardiologist, but I can't imagine the day would ever come where using

00:36:47.840 a stethoscope, I could detect, you know, a low EF, like to the tone of... Yeah. No, no. Right.

00:36:53.140 Yeah. I mean, unless there's a huge murmur. No, no. It's really the regurgitation that they're

00:36:57.440 hearing as a heart murmur. And that was most of the dogs that we ended up having to exclude,

00:37:02.580 we had to exclude because they had a pre-existing value of regurgitation that came out from the

00:37:09.060 echocardiogram. But this was actually a discussion that we had to have with the cardiologist. You

00:37:14.360 know, the cardiologist initially went into the study with the feeling that if there's any regurgitation

00:37:21.020 that's beyond trivial, that that's heart disease. And so when we started to see dog after dog show up

00:37:27.440 with this level of regurgitation, we had to have this discussion. What's normal aging versus disease?

00:37:32.580 Right. And so just because of the way that the protocol was written, we ended up having to

00:37:37.300 exclude about 20% of dogs because they had underlying heart disease. So we had 40 dogs come

00:37:42.340 into the study for their first exam. That was our target number. We ended up having 24 go all the

00:37:47.140 way through the study. Three groups, two doses plus a placebo. Yeah. And they ended up not being evenly

00:37:52.620 distributed. And in part, this was because, well, in part it was because it was my first clinical trial

00:37:57.780 that I'd ever done. And I didn't plan for the fact that we would have to exclude as many dogs as we

00:38:02.420 did. So the way it ended up was we had eight, eight placebo, 11 high dose and five medium lower dose.

00:38:10.800 Yeah. Right. That went all the way through. We only had one dog leave the study. And that was because

00:38:15.280 the owner just stopped giving their dog the medication. And we had one dog where the owner

00:38:20.180 gave their dog the wrong amount of the medication, which ironically enough, he was a physician low.

00:38:26.120 So the dog was randomized into the high rapamycin group and ended up getting one quarter of the

00:38:32.060 expected dose. So other than that, there was fantastic compliance. All of the owners did what

00:38:37.460 they were supposed to, came in for their exams. So the main outcomes of this study were one, there was

00:38:42.980 no evidence for increased side effects. So the owners filled out weekly surveys that there was a long

00:38:50.100 list of... The dogs didn't fill them out. Well, you never know. There was a long list of did your dog

00:38:55.500 experience any of these things? And then there were a couple of just open-ended questions. Do you feel

00:39:00.840 like you observed any positive changes or negative changes in health? Did the group in Tennessee see

00:39:06.200 the apthos ulcers in their dogs? No. Because that seems to be... I remember as a resident when we would

00:39:12.260 give rapamycin to the kidney transplant patients. I mean, the biggest complaint by far was those apthos

00:39:17.700 ulcers. Yeah. And we didn't, again, in hindsight, we probably should have done a better job of

00:39:23.040 looking, but we did not have any evidence that that was happening. The only thing that initially I

00:39:29.440 thought that maybe we were detecting that was because we had several owners, and this was while

00:39:36.100 the study was still blinded. So I didn't know which dogs were which at this point. We had several

00:39:40.920 owners report that their dog was drinking a lot more water. And a couple of them actually came in

00:39:47.680 for urinalysis and stuff like that. And so I thought maybe that could be like a dog's response.

00:39:53.520 If they have sores on the inside of their mouth and they're uncomfortable, maybe that would be the

00:39:57.540 sort of the canine equivalent of how they would respond to that. As it turns out, though, when we

00:40:02.120 unblinded the study, that reported observation of increased water consumption was equally spread

00:40:07.460 between the placebo and rapamycin groups. So I don't know if that's happening in dogs. That's

00:40:11.580 something we'll look at in the next phase. So again, for all of the side effects, though,

00:40:16.240 that we surveyed owners on no difference between treated and untreated, the blood chemistry showed

00:40:23.160 no significant changes with rapamycin, which was actually a little bit surprising.

00:40:27.320 Did you do a glucose tolerance test?

00:40:29.020 We did not do a glucose tolerance test, in part because we wanted to keep the number of assays that

00:40:36.700 we were asking the owners to subject their dogs to as small as possible and make it as non-invasive as

00:40:42.560 possible for the animals. We did get blood chemistry at week, you know, before randomization at week

00:40:48.120 three and at week 11. So within one week of coming out of the study. And we saw improvements in heart

00:40:53.740 function. Now I will say it's a small cohort. They were on sort of a borderline of statistical

00:40:58.920 significance. So two of the three measures that we had as our primary endpoints, the three measures

00:41:04.120 were ejection fraction, fractional shortening, and E to A ratio. And again, that was just coming

00:41:08.200 directly from the mouse studies. Two of those three were statistically significant. One was

00:41:12.820 p-value of 0.06.

00:41:14.520 Yeah, but I mean, you must have been underpowered on virtually anything and everything.

00:41:19.020 It's amazing you saw significance.

00:41:20.520 I agree. I agree. Yeah. So I think the way I view this is it's about the most positive outcome

00:41:26.560 that we could have hoped for. Clearly needs to be replicated, but at least the changes are going in

00:41:32.740 the right direction. And an interesting couple of interesting things when you actually look at the

00:41:37.480 heart data, it very much looks like the dogs on rapamycin that got the biggest benefits were the

00:41:44.180 ones that started with the lowest function, which is not surprising, but that also is encouraging

00:41:49.220 because that's, that's kind of what you'd expect, right? The dogs that have undergone a greater age

00:41:53.960 related decline are likely to be the ones that are going to get the biggest benefit from a treatment

00:41:58.820 that's actually affecting that. So that was really encouraging. And then we actually had one

00:42:03.400 Doberman Pinscher in the study. And this turned out to be interesting because Doberman Pinschers as a

00:42:10.460 breed are highly prone to dilated cardiomyopathy. Something like 60, 65% of Doberman Pinschers will

00:42:18.260 develop dilated cardiomyopathy as they get older. I wasn't really aware of this literature going into

00:42:23.360 the study. But it turns out that many Doberman Pinscher owners will actually give their dogs

00:42:30.960 echocardiograms or electrocardiograms routinely as they're getting older to try to detect dilated

00:42:38.620 cardiomyopathy as early as possible. And this, the owner of this dog didn't tell us this before she

00:42:44.080 came into the study, had actually been aware of this and was giving her dog echocardiograms before

00:42:49.160 coming into the study. That dog had low cardiac function, but it was not yet to the point where it

00:42:55.540 was clinically diagnosed as dilated cardiomyopathy. So our cardiologist in the study, also not knowing

00:43:01.240 the dog's history of having prior echoes, did not flag the dog as needing to be excluded. So the dog

00:43:07.940 was randomized. It just happened to be randomized into the higher rapamycin dose group. And its function

00:43:13.940 was, that was one of the dogs where we saw the largest improvement in function. What is interesting

00:43:18.700 about that is the owner then after the study was over continued to get echocardiograms every three

00:43:24.380 months and has shared that information with us. And so it's really, it's an N of one, but it's a

00:43:28.760 really fascinating sort of case study because you can see the dog's cardiac function declining. Then the

00:43:34.620 dog comes into the study, gets rapamycin, it shoots up. It's quite a dramatic increase. And then within

00:43:40.140 about... What was the increase in EF? Do you remember? I don't remember. I don't want to say the exact

00:43:45.600 numbers because I don't remember off the top of my head. It sounds like it was a big deal. It was from the

00:43:48.640 borderline of being a cult dilated cardiomyopathy up at least 10, 10%. 10%. 10 absolute percent.

00:43:55.420 That's right. That's enormous. Yeah. Well back into the normal range. Yeah. And has that patient or that

00:44:01.120 patient, has that woman shared with you what the resulting decline in EF has been since the trial?

00:44:06.880 So about, I've got the data out to about six months and we're just now trying to reconnect with her to

00:44:11.500 see if she has additional data that she'll share with us. By about six months out, the ejection

00:44:17.560 fraction and fractional shortening were back right at the point when the dog came into the study. And

00:44:23.040 at that point, her cardiologist diagnosed the dog as a cult DCM. So clearly going down the path to

00:44:29.500 dilated cardiomyopathy. So the million dollar question in a study like that, or in a case like that is,

00:44:35.520 if you had to guess what would be the ideal way to take care of that dog to delay the onset of

00:44:44.400 cardiomyopathy as long as possible? Would it be just keep this dog on that dose three times a week

00:44:50.400 in perpetuity? Would it be give the dog a 10 week holiday, 10 on, 10 off, 10 on, 10 off? Right.

00:44:56.320 So it's clearly 10 on every six months is just a seesaw. If everything's working the way that we think

00:45:03.380 it is. Yeah. My guess is that the default there would be to keep the dog on the drug unless you

00:45:08.400 start to see side effects. Right. So, so continue to monitor by echoes every three months. And unless

00:45:13.820 you see side effects or unless you see something else that makes you worried that rapamycin is,

00:45:18.360 is having a negative effect, just keep the dog on, on the drug. But you know, it's an unknown as to

00:45:24.640 whether we would eventually see side effects at that dose because, you know, as I said, the only data

00:45:29.700 that I know of is that university of Tennessee group where they did have some dogs that survived

00:45:35.300 more than a year and continued to take the drug. Those dogs, you know, as I said, had had hemangiosarcoma

00:45:41.080 of the spleen. They'd had surgeries, you know, they were very sick coming into that study. And so I don't

00:45:45.980 know, even if there were mild side effects that you would really be able to tease that apart from

00:45:51.140 everything else that's going on with those dogs. There are other larger mammal studies that are going on,

00:45:56.520 correct? With rapamycin specifically. So there are, well, so, so there are research studies in the

00:46:02.780 context of aging in marmosets. They're very small and, and very early, but, but those are being done

00:46:08.060 at the University of San Antonio, yeah, Texas, San Antonio, the Bar Shop Institute. And then there are

00:46:14.740 a variety of cancer studies in dogs with rapamycin. So there's a large study of rapamycin for osteosarcoma

00:46:23.400 in dogs. And then there are a few smaller clinical trials, but I don't know of any other

00:46:28.460 large animal studies in the context of aging. Anti-aging, yeah. Yeah. Now I want to obviously

00:46:35.860 come back to the anti-aging in the dogs, but on the cancer topic, there are some data that are

00:46:41.860 actually suggesting that the increase in autophagy that one might see with rapamycin, which one would

00:46:50.480 expect to see, at least with mTORC1 inhibition, might be paradoxically not ideal in the active

00:46:58.140 setting of cancer. Right. So I guess my first question is, are you swayed by those data? And if

00:47:04.900 so, what would be the teleologic explanation? Yeah. So I'll say I'm not swayed much. I think it's,

00:47:11.920 I think that those kinds of data are important to be aware of. I think one of the real challenges

00:47:17.020 in the autophagy field is that not everybody- Besides the fact that we can't measure it easily?

00:47:21.020 I was going to be one of them. Yeah. Beside the fact that we can't, we don't really know how to

00:47:25.760 measure it. I think as a community, we really don't know what we mean when we say that autophagy

00:47:31.520 is increased or decreased. Different people use different markers or measures for autophagy.

00:47:37.960 So my view is that one of the things that can happen, not just for cancers, but for lots of

00:47:44.000 different pathologies is one of the ways that cells try to deal with many different forms of

00:47:53.440 stress, in particular protein misfolding, but also other forms of stress is to turn up autophagy.

00:47:58.940 So I think that autophagy, some markers of autophagy going up can be a response to a pathological

00:48:07.440 condition. Also, what often happens is that response of turning up autophagy does not lead

00:48:14.160 to productive autophagy. So you actually get an accumulation of autophagosomes because they don't

00:48:19.160 ever make it all the way through the process. And so depending on how you're measuring autophagy,

00:48:23.380 what you really might be detecting is a block at the late stages of autophagy. And what you're seeing

00:48:29.180 is- You're measuring a backlog.

00:48:30.640 That's right. So it's not necessarily the case that more autophagy is bad in that context. It's

00:48:36.900 the failure to actually bring it through to completion. And again, I don't have a lot of

00:48:41.500 data to support this. My intuition is that at least for some diseases, one of the things that

00:48:48.180 rapamycin does, and I don't really understand the molecular biology here, it seems to alleviate that

00:48:54.560 block. So you actually get productive autophagy working again. And again, I don't know exactly how

00:48:59.260 that works, but that's what it seems to me is happening. And so we have seen some evidence for

00:49:03.800 this in mitochondrial disease in the brain, where if we look in the brain, we can see these sort of

00:49:09.460 massive autophagosomes that are trying to digest mitochondria that can't do it while the disease

00:49:15.640 is progressing. Somehow rapamycin fixes that. So I think you have to be cautious in interpreting an

00:49:21.900 increase in autophagic markers in a disease as necessarily meaning that increased autophagy is

00:49:29.140 causing or contributing to that disease. And it could be the case that depending on how you

00:49:33.340 activate autophagy, it could be detrimental or it could be beneficial. And so there's really two

00:49:39.240 different questions. The first would be, if you take a patient with cancer and you inhibit mTOR,

00:49:47.480 is it not helpful because the tumor has already evolved so much to be outside of mTOR's purview?

00:49:58.680 Or is it, it's actually harmful? And that's of course, separate from the option that it could

00:50:04.000 be helpful. Right. So my understanding of the clinical and the literature in humans is that for

00:50:11.360 most cancers, once it's reached the point of diagnosis, that rapamycin is disappointing in

00:50:19.540 its effectiveness. It's not particularly effective. That's not true for all cancers, but for most

00:50:23.760 cancers, it has not been as effective as you might expect, given that we know that activation of mTOR

00:50:29.760 is common when you get high proliferation. And then turning down mTOR should stop that proliferation.

00:50:36.780 Turn off a proliferative cell. Yeah. So I think, I think you're probably right that at least part of

00:50:40.660 the story is that one of the steps in the progression to cancer is evolving to ignore

00:50:46.480 that signal. The break. Yeah. Of turning down mTOR. So rapamycin may not be effective there. I think

00:50:52.120 it's a complicated system though, because the effects of rapamycin on the immune system

00:50:56.760 could have beneficial effects in terms of cancer or detrimental effects. So we know that immune

00:51:03.060 surveillance is probably the most important anti-cancer mechanism, or certainly one of the

00:51:07.660 most important anti-cancer mechanisms. And we know that immune function goes down with age. That's

00:51:11.920 probably one of the reasons why most cancers are age-related. So if you can boost age-related

00:51:17.500 immune function with rapamycin, enhance immune surveillance, that's going to have a potent anti-cancer

00:51:22.960 mechanism. And again, this is my guess. My guess is that's why we see in the studies in mice that

00:51:30.460 cancers are pushed back during aging by rapamycin. On the other hand, if the dose of rapamycin is high

00:51:39.240 enough that you're actually inhibiting immune function, that could be, that could promote

00:51:43.840 cancers. That could amplify. Yeah. And there's not a lot of data yet. So we did one study in my lab where

00:51:48.620 we gave mice, I think it's the highest dose that's ever been given in the context of an aging study.

00:51:53.360 This was a daily injection of eight milligrams per kilogram. So that's, we call it the party dose.

00:51:58.340 Yeah. Right. Right. And so this was a study where we only gave the mice rapamycin for three months.

00:52:04.060 So this was from 20 to 23 months. And then we stopped the treatment. And what was interesting there

00:52:09.280 was we got completely different effects in male mice versus female mice. The male mice lived 60% longer

00:52:16.200 after the end of treatment. They had better muscle function. They got less cancer. The female mice

00:52:22.600 had no difference in lifespan. The mice that got rapamycin or didn't get rapamycin, but they died

00:52:29.140 with, I want to say from, but it's hard to say for sure what a mouse dies from. They died with

00:52:34.680 very different types of cancers. So the female mice that had gotten this high dose of rapamycin for three

00:52:39.740 months, all had aggressive hematopoietic cancers. Whereas about, I think it was about 30 or 40% of the

00:52:48.640 vehicle treated mice. So in black six, that's not an uncommon cancer to get. But none of the rapamycin

00:52:54.400 treated mice had non-hematopoietic cancers. Whereas like 60% of the mice that didn't get rapamycin.

00:53:00.480 Now the 2009 study that kicked all this off actually showed a greater survival benefit in the female mice,

00:53:05.560 didn't it?

00:53:05.860 That's right. So I think, and again, this is a guess because I don't actually have the data to

00:53:10.100 back it up. My guess is that because we pushed the dose so high, we might've actually taken it

00:53:15.840 too far in the female. So one school of thought is that female mice, at least, we don't know if this 0.66

00:53:21.660 is true in any other organism. Female mice are more sensitive to rapamycin. And that could either be 0.97

00:53:27.220 that they don't clear the drug as quickly or that for whatever reason in female mice,

00:53:33.400 the same amount of rapamycin has a greater mTOR inhibitory effect. But that's one school of

00:53:39.440 thought. And I kind of think that's right. So at lower doses of the drug, you see a bigger lifespan

00:53:43.940 benefit in females than in males.

00:53:45.860 Did you repeat that experiment at like four megs per kg or something different?

00:53:48.840 No, we haven't. We haven't. We should. So we did do...

00:53:52.100 We just need an infinite pool of money to do all of these, like just answer all these questions.

00:53:56.240 At least to figure out the most important questions. Yeah.

00:53:58.140 And I think the dose response is really important. We did do a lower dose for three months as

00:54:02.980 well. And there we saw increases in lifespan in both males and females, roughly the same

00:54:07.280 magnitude. So it was that dose was nine times higher than what the ITP tested.

00:54:11.820 Wow.

00:54:12.140 So one of the things that's interesting though, is as you go higher in dose, so three times

00:54:17.040 higher than what they originally tested, the females still live a little bit longer, but 0.78

00:54:21.620 the difference between males and females, the gap has closed quite a bit. So I think that females,

00:54:26.620 for whatever reason, at a given concentration of rapamycin are just more affected by that

00:54:31.680 amount of the drug. And I think what we did in our high dose study is we just pushed it

00:54:37.640 a little too far. We pushed it to the point where rapamycin did something probably to the

00:54:43.540 immune system that allowed these immune cancers to escape surveillance or become hyperproliferative.

00:54:50.420 And again, I'm not a cancer biologist. I'm not an immunologist. So I don't have a good

00:54:55.900 feel for what the mechanism is. I can tell you what the observation is. And that's that all of

00:55:00.220 those animals had aggressive hematopoietic cancers when they got this three months of rapamycin.

00:55:06.240 Just out of curiosity, more B-cell or T-cell? Do you recall?

00:55:09.220 I don't recall. It's in the paper. We could look it up.

00:55:12.060 Because there's an opportunity here to do the reverse, right? I mean, there's an opportunity 0.59

00:55:14.540 to take, right now we're seeing just an unbelievable amount of activity in adoptive cell therapy.

00:55:19.240 And, or even when you just talk about like checkpoint inhibitors and things like that,

00:55:23.680 like it makes you wonder, are there ways to make these things better? Maybe the checkpoint's the

00:55:27.440 wrong example because you might get more autoimmunity. But certainly when you talk about

00:55:31.180 adoptive cell therapy, anything that could boost either, you know, CD8 function or inhibit the regs

00:55:38.240 or something, there might be ways. Like it almost makes you wonder if using rapamycin in a different

00:55:43.700 manner in combination with immune-based therapies might make more sense.

00:55:48.240 Yeah, no, I think there's a lot that could be done there for sure. Part of the reason why we

00:55:51.740 haven't explored this in more detail, well, one reason is, again, as I said, I'm not a cancer

00:55:56.100 biologist, so it's not, that's not the thing I'm most interested in. I think it's really interesting

00:55:59.800 biology, but it's not the thing I'm most interested in. But I also feel like because the dose that we

00:56:04.540 gave was so high that, again, thinking translationally about rapamycin as a drug in the context of aging,

00:56:11.820 my feeling is that what we've uncovered here is not going to be relevant at the doses that we would

00:56:19.180 think about giving to dogs. So that's why I haven't really spent a lot of my time trying to figure out

00:56:24.780 what's going on there. But I think certainly in the context of cancer immune therapies, I think we do

00:56:30.640 need to think a little bit more about how effective those kinds of therapies are going to be in the

00:56:36.960 elderly. And maybe something like rapamycin could help, could actually enhance the ability of those

00:56:43.440 therapies. I mean, this question you posed when David Sabatini, Tim Ferriss, and Nav Chandel and I

00:56:48.480 were in Easter Island a year ago, over a year ago, this might have been our favorite mealtime discussion,

00:56:54.740 which is what best explains the increase in cancer incidence with age? In other words,

00:57:02.480 would the primary driver be the reduction in immune surveillance or the length of time to

00:57:08.120 accumulate mutations or the frequency of mutations? Like, I mean, it's not an obvious answer. And I

00:57:13.760 don't think it has to be just one. I think all those things are working together. Yeah. Yeah. I

00:57:18.440 certainly over the last few years have come to think that the decline in immune function is,

00:57:23.960 it's certainly more important than I had initially thought. That's my... I mean, I secretly want that to be

00:57:29.300 the biggest driver because I think we have a better chance to control that than some of the other

00:57:33.660 ones. And I think it probably is. That would be my guess. And I also think it kind of makes sense that

00:57:38.640 if you have an immune system that's functioning the way it's supposed to, you can actually deal with

00:57:45.100 the mutation accumulation because your immune system is going to clear those before they become

00:57:50.480 problems. So now let's go back to the anti-aging thesis, right? Which is we're going to take

00:57:56.880 healthy dogs, eventually healthy people. We want to reduce the rate of decline is probably the best

00:58:05.620 way to think about it, right? So we have a deterioration in organ function. That's the way

00:58:08.720 I used to think about it. Okay. All right. One of the things that's been surprising to me again over

00:58:13.100 the last few years is the different ways that rapamycin not only seems to delay the decline,

00:58:19.380 but it seems to make things better. There clearly seems to be in at least some organs a rejuvenating

00:58:26.800 function. And I suspect that's mostly stem cell mediated, but again, the mechanisms haven't been

00:58:31.620 worked out yet. So we've already talked about immune function. You can take an old immune system

00:58:35.600 and make it work more like a young immune system. We've talked about cardiac function. You can take

00:58:39.760 an old heart, you make it work more like a young heart. There's some evidence from David's lab

00:58:44.680 that intestinal stem cells can be rejuvenated by rapamycin. We've recently published that alveolar

00:58:51.000 bone levels. So in the mouth, the bone around the teeth can be rejuvenated back to a more youthful

00:58:56.540 level by short-term treatment with rapamycin. So there are now multiple different places in the

00:59:03.220 body, at least in mice, where you actually see functional improvements back to a more youthful state.

00:59:10.080 And so I don't think rapamycin is going to do that for everything, but at least for

00:59:14.180 tissues and organs where stem cell senescence plays a big role, I suspect that rapamycin can have

00:59:21.140 not just an effect on delaying declines, but actually bring things partially back towards a

00:59:26.820 more youthful functional state. What's the best available evidence for that centrally in the CNS?

00:59:32.780 That's a good question. So there are studies on cognitive function in mice showing that you can

00:59:41.340 improve cognitive function in aged animals. I believe at least one of those started the treatment

00:59:48.040 late in life and saw improvements in cognitive function. And then in all of the major Alzheimer's

00:59:54.360 disease mouse models, you can, the literature is a little bit mixed. There's at least evidence that

00:59:59.400 you can wait until the pathology of the disease has set in. You see the A beta accumulated. You see the

01:00:05.760 functional deficits in terms of cognitive function. You can start the treatment and you can improve

01:00:09.780 things. And on a post-mortem, are you seeing an actual reduction in animal? Yeah, you are.

01:00:15.820 So I mean, we haven't done this. This is the work of several other labs in the AD area. Yes. So again,

01:00:21.640 the data is a little bit mixed. So there are a couple of papers out there where they, they see that

01:00:26.300 you can get really robust benefits if you start rapamycin treatment early, but they didn't see

01:00:31.520 benefits in the AD models when you started late. And then there are studies that, that did see

01:00:36.640 declines in aggregation, increased autophagy and functional improvements.

01:00:41.620 I mean, it would be staggering if you could take an animal and ultimately, of course, a human

01:00:46.760 who's already accumulating AB and tau. Yeah.

01:00:50.580 And even just halting that is a big deal. There's one drug in, you know, that is approved.

01:00:55.900 Now you're going right into my biggest pet peeve right now, which is why there hasn't been or isn't

01:01:01.280 a current rapamycin trial for Alzheimer's disease. But I think you're right. And I, and again,

01:01:05.520 three years ago, if somebody had asked me, will rapamycin, is it likely to have any benefit in

01:01:10.620 somebody who's been diagnosed with AD? I would have probably said, no, I come around to thinking

01:01:15.800 that there's at least a reasonable chance that it can not just halt progression, but it could

01:01:21.420 actually make, well, especially if you, you know, one of our colleagues is a, is a neurologist here

01:01:26.300 named Richard Isaacson, and he runs the largest Alzheimer's prevention clinic in the country at

01:01:30.400 Cornell. You know, Richard's thesis, which I think makes a ton of sense is, again, I think other

01:01:35.720 people share this view is you want to catch people while they just have the first signs of mild

01:01:40.140 cognitive impairment. Sure. Absolutely. And your ability to actually impact them is enormous. And

01:01:45.220 so the question is, why aren't those people being considered for clinical trials when we all already

01:01:51.180 know that these other agents aren't really doing anything? Right. Yeah. I agree. I think that if I was

01:01:55.880 going to design a clinical trial for Alzheimer's disease or dementia with rapamycin that I thought

01:02:02.940 had the best chance of success, that would be the target population. Those trials are harder to do in

01:02:08.180 some ways because they're longer, right? And not everybody moves from MCI to full-blown AD at the same

01:02:14.960 rate. And we don't have, at least my understanding is we still really don't have great predictive biomarkers of

01:02:21.160 how fast that's going to happen in an individual. But that would be the, that would be the study that

01:02:25.940 would have the best chance of working. Having said that, I still think there's a decent chance

01:02:30.420 in somebody who's, you know, already gotten to the point where, where they will be diagnosed as

01:02:37.120 having Alzheimer's disease, serious functional deficits. There's a chance that rapamycin could make

01:02:43.140 things better. Now I get the practical reasons for why people don't want to try a risky clinical trial.

01:02:48.040 It's expensive. If you fail, you know, you're, then your drug gets a bad reputation, all of that.

01:02:52.720 So I understand why people are hesitant to do that trial. I think there's actually a pretty good

01:02:57.280 chance it would help people. Especially if you can combine it. I mean, I think, you know,

01:03:00.940 Richard often talks about, and others do as well, that one of our failures in Alzheimer's is we

01:03:06.520 consider it a single disease. I agree. It's as naive as saying, John has cancer. Oh, well, gosh,

01:03:11.980 that's the end of his life. Well, don't you want to know what kind of cancer first? Or maybe what

01:03:15.940 mutation. So similarly, but broadly speaking, and this is a gross oversimplification, if you

01:03:20.120 consider the metabolic version of Alzheimer's disease, the vascular version of Alzheimer's

01:03:25.360 disease, and then the sort of toxin clearance impaired version of the disease, to me, I'm

01:03:31.160 generally most optimistic about the metabolic one. So the variant of this disease that seems to be

01:03:36.300 mostly due to a failure of energy metabolism in the brain, that also strikes me as the one that's

01:03:42.220 most amenable to systemic therapies as well. If you improve insulin sensitivity, if you improve

01:03:49.320 glucose disposal, if you reduce hypercortisolemia, if you can actually, through nutrition, exercise,

01:03:55.620 sleep, a number of other things start to modulate that. That strikes me as the one where you want to

01:04:00.580 at least take your first shot at adding rapa. I think you're probably right. Although again,

01:04:04.240 with rapamycin, because we know that it is effective at, for example, turning up autophagy,

01:04:09.440 it might also be... It might work in the harder ones. Yeah, that's right. But I agree. I tend to

01:04:13.780 agree with you. I think that's a good point. I also want to add that I actually think the biggest

01:04:18.760 problem with the way that the scientific community has thought about Alzheimer's disease, aside from

01:04:24.220 considering it to be one disease, is not really recognizing that it's a disease of aging. I really

01:04:31.500 think that one of the reasons why the preclinical research has been disappointing at developing

01:04:38.860 therapies for Alzheimer's disease is because very rarely have people approach that from the

01:04:45.740 perspective that this is a disease of aging. And so something that can affect the mouse models of

01:04:52.000 Alzheimer's disease when we create this disease in young mice may not work the same way in an aged

01:04:59.260 person or an aged animal. And that's one of the things that also makes me optimistic about rapamycin,

01:05:04.660 is we already know that it hits the hallmarks of aging. And it also seems to be effective in

01:05:11.000 these Alzheimer's disease models. So that makes me think that it's acting at a sort of a more

01:05:16.020 fundamental level to target the molecular causes of this disease. Well, especially, look, I mean,

01:05:21.340 if you can regenerate a cardiac myocyte from its stem cell, it's not an impossible thought that you

01:05:27.860 could regenerate neurologic stem cells. Absolutely. And this is, again, an area where... Which,

01:05:32.060 I mean, 20 years ago, we would have said that's impossible. It's metaphysically impossible. Right.

01:05:35.960 Going back to the dogs for a moment. I, again, I'm not familiar with dog literature. What is the

01:05:41.100 fasting literature look like in dogs? There's not a lot that I'm aware of either. And I think that

01:05:46.740 kind of makes sense. So, you know, first of all, you have to differentiate the literature in laboratory

01:05:51.900 colonies from companion dogs. Companion dogs, right. There's probably not much in companion dogs.

01:05:57.380 I don't know of true fasting experiments in like beagle colonies. So we can't use like the fasted 0.88

01:06:05.400 versions as proxies for what we would hope to see on the rapa dogs, which is the way they sort of did

01:06:12.100 it in mice, right? They sort of said, well, you know, we know that if you calorically restrict this

01:06:16.100 mouse, this strain of mouse under this degree of caloric restriction can expect as much of a

01:06:20.240 longevity boost. And oh, lo and behold, rapamycin is probably even better than that.

01:06:24.200 Right. Although from a metabolic perspective, I think it's still unclear, even in mice,

01:06:30.000 whether rapamycin and caloric restriction are working through the same mechanism. And this is,

01:06:35.060 this is, this is actually another area where there haven't been a lot of good experiments done.

01:06:40.140 So there's a portion of the field that argues strongly that the caloric restriction and rapamycin

01:06:46.000 are completely different, which to my mind is absurd. I mean, we know that one of the main

01:06:51.740 things that caloric restriction does is it inhibits mTOR. Right. And we know that rapamycin inhibits

01:06:56.400 mTOR. So they're not fundamentally different. So I think they are overlapping, but distinct. Not

01:07:01.640 everything that caloric restriction does is going to be mimicked by rapamycin and vice versa. Having

01:07:07.360 said that, when you look at the gene expression profile or the metabolic profile, they don't look

01:07:12.760 all that similar, at least at the low doses of rapamycin. And so I think it's, it's a little bit

01:07:18.460 unclear. You're right. Lifespan, they both extend lifespan. Rapamycin might actually extend lifespan

01:07:23.000 across a broader genetic background than caloric restriction. But when you get beyond that, I think

01:07:29.100 it's still an unknown whether there are, you know, say, are there metabolic signatures that are common

01:07:33.680 to both that might therefore be more likely to be causal. I think, I think we don't know the answer.

01:07:38.480 So another, what you're really getting at is probably the next thought I had subconsciously,

01:07:42.780 which is, could we use calorically restricted animals to develop a metabolic signature as the

01:07:48.600 gold standard for how we would then titrate rapamycin both in dose and frequency? I don't

01:07:54.580 know of any evidence to support that. That would be a stretch. Yeah. I mean, it might be possible,

01:07:59.080 but there just isn't much good data out there at this point. I do feel like if we want to put the

01:08:03.940 resources into trying to identify predictive signatures, which I absolutely think we should

01:08:09.700 do. I think the, the metabolome is probably the place to go and the serum metabolome makes a lot

01:08:16.140 of sense. So that would be where I would look and nobody's really done it well. Part of the problem

01:08:21.400 is the data that's out there for rapamycin is again, all at that very low dose that the ITP tested

01:08:27.920 originally. We know that suboptimal for lifespan. It's quite likely that the effects that the lowest

01:08:33.740 dose of rapamycin are having on metabolism are going to be relatively modest compared to higher doses.

01:08:39.080 So you, the changes might be there, but they might be so subtle that you're not going to detect them

01:08:44.520 in a sort of high throughput screening approach. So, so I think there's a lot we don't know,

01:08:49.900 but that would be where I would put my effort in trying to identify serum metabolomic signature

01:08:56.500 of rapamycin that we could then correlate with the effects on function in a variety of different

01:09:03.160 tissues as well as life. Yeah. Because I mean, I think as wonderful as it would be to just wave

01:09:07.320 magic wands and have biopsies of tissue, it's just not going to be, I mean, we're not going to take

01:09:11.160 cardiac biopsies. Yeah. Right. So you, if you're looking for predictive signatures, you can do those

01:09:16.760 experiments initially in mice, but you really have to think about, is this something that's reasonable

01:09:21.520 to do in a person or a dog? And have you looked at the gut biome in the dog at all? A little bit.

01:09:27.380 So, so we did a study in mice. This was that, that study I referred to previously where we treated them

01:09:32.660 for three months with either the high, high dose or the high, but not as high dose. Yeah. And we did

01:09:37.660 look at the fecal microbiome and we saw quite dramatic changes there, some of which are interesting.

01:09:43.200 And so that's, that's an area that we would love to pursue. I've tried to get a grant to do that and

01:09:49.240 so far haven't been successful through NIH. But it hasn't been done in dogs yet. Has not been done in

01:09:53.760 dogs. So we have a little bit of preliminary data from our phase one study, and we're seeing changes

01:09:59.620 in the microbiome there as well. So far, the data that we've got in dogs is not the sort of

01:10:05.040 comprehensive metagenomic approach where we sequence everything that's there. It's more of a standard

01:10:09.480 veterinary approach where they, they have sort of broad classifications of different types of

01:10:14.240 microbes. And there are definitely changes with rapamycin. It looks, it's very early, but it looks as

01:10:20.140 though, at least in dogs that have a bad microbiome, a dysbiosis in their microbiome, which can be

01:10:26.740 defined clinically from this test. There were two dogs that we've looked at that started out with a

01:10:32.360 bad microbiome. They were in the rapamycin group. They were better by the end of the study. Obviously

01:10:37.440 don't know if that's meaningful or not. So, so there are changes in the dogs, but it's really too early for us

01:10:42.240 to know whether they look like the mice and, or whether that is going to either be predictive or

01:10:48.200 potentially play a role in some of the effects of rapamycin. It is definitely the case though,

01:10:53.600 in both dogs and mice, and I'm sure this will be true in people as well, that rapamycin has a large

01:10:59.580 effect on the composition of the gut microbiome. I suspect that will also be true for other

01:11:04.800 compartments in the body, like the oral microbiome, maybe even the skin microbiome.

01:11:09.100 Nobody's looked yet. So that there's a lot to be done there. I also don't necessarily think that's

01:11:13.160 going to be unique to rapamycin. I think that lots of drugs that we take.

01:11:16.420 Yeah, food will change that. And absolutely diet.

01:11:18.680 And it's not clear if that's the effect or the cause of it.

01:11:21.440 Yeah. So one of the reasons why I think that there's reason to think that at least some of

01:11:25.000 the changes in the microbiome could be causal for some of the effects of rapamycin is that

01:11:28.700 one of the things we saw in our mouse study was a pretty profound increase in a bacterium

01:11:36.600 called segmented filamentous bacteria or SFB in the rapamycin treated mice. It turns out if you look

01:11:43.020 in the literature, there are links between SFB and diabetes and obesity, and also between SFB and

01:11:51.460 T helper cell maturation. So it could be the case that changes that rapamycin is having on this

01:11:58.200 specific bacterium, as well as other bacteria, are then having effects both on potentially nutrient

01:12:04.820 utilization and uptake, but also direct effects on, say, immune function. I mean,

01:12:09.920 the intestine is an important immune compartment. The bacteria are physically right there. These SFB

01:12:17.280 actually form filaments directly associated with the intestinal cells. So it certainly could be the

01:12:23.220 case that there are signals being sent back and forth that rapamycin is modifying by, through

01:12:29.280 modification of the composition of the microbiome, and that that's affecting immune function,

01:12:33.820 adiposity, all sorts of different possibilities. So like I said, that's something we're really

01:12:37.980 interested in testing. So there's like an infinite number of things I want to know, right? There's

01:12:41.980 this company out there that's looking at duodenal ablation to ameliorate diabetes. And the data are

01:12:47.000 actually really interesting. So interesting, in fact, that when I first saw them, I thought this

01:12:50.740 has got to be nonsense. What's not clear is the durability and the economics of it. But just from

01:12:54.820 a scientific standpoint, they're doing these ablations of the duodenal mucosa, and they're seeing

01:13:00.860 like immediate step function changes in insulin sensitivity, arguing, in fact, that this may be

01:13:06.580 the mechanism by which the Roux-en-Y gastric bypass is ameliorating type 2 diabetes. It's basically

01:13:11.980 taking this dysfunctional duodenum out of the loop and just saying, we're going jejunum to jejunum,

01:13:18.260 or, you know, stomach to jejunum to jejunum, basically.

01:13:20.800 It's really interesting. So this is another area that I've gotten interested in, and we haven't

01:13:24.940 really dove into it yet much. But I think there's clearly literature growing in all of the model

01:13:33.540 organisms that with aging, there is a decline in intestinal barrier function, and that at least

01:13:40.920 in flies, that seems to be causal for death. So in other words, it's strongly correlated, and there

01:13:46.940 are ways, every way that extends lifespan also seems to improve this intestinal barrier dysfunction.

01:13:52.540 And I've been thinking, and we have, again, a little bit of data suggesting that rapamycin might

01:13:57.960 actually have an effect on this age-related decline in intestinal barrier function. And

01:14:01.920 why would a loss of intestinal barrier function be important? Well, I mean, one thing that happens,

01:14:06.520 we know this happens, is that you tend to see an increased level of microbial proteins and DNA

01:14:15.080 in the circulatory system with age. And that causes an inflammatory response that may contribute to

01:14:22.140 the sort of systemic increase in inflammation that we see during aging. So the loss of intestinal

01:14:27.940 barrier function may actually drive, to some extent, the inflammation, the increase in inflammation

01:14:33.740 with aging, or at least contribute to it. And so anything that you can do that is going to improve that

01:14:39.280 will potentially have an effect on systemic inflammation. And that, again, could be another way

01:14:44.140 that rapamycin is sort of impacting the entire body in addition to the effects of rapamycin when it gets

01:14:52.160 to a cell and inhibits mTOR in that cell. And whether that's through changes in the microbiome or changes

01:14:58.680 in intestinal stem cell, like Sabatini has shown, right, there's lots of possible ways that it could

01:15:02.940 be working. But I do believe that this decline in intestinal barrier function with age probably is

01:15:09.480 contributing to this sort of increase in systemic inflammation that we see during aging. And it's

01:15:14.740 very clear that that happens in people and in non-human primates as well during aging.

01:15:19.520 Now, talking about people again, I don't hold out much hope that there's going to be

01:15:25.140 an anti-aging trial in humans using rapamycin or a rapalog. You know, watching how much difficulty it is

01:15:33.620 to even try to get that study done with metformin, which, I mean, is about the most inert, you know,

01:15:38.360 it's like, you might as well just do it with drinking water. We're going to randomize you to

01:15:41.840 seltzer versus flat water here. So the real question is, for the people who want to be on

01:15:45.960 the tip of the spear, you're not going to have a gold-plated stamped RCT. You're going to have to

01:15:52.100 triangulate from everything else. Many roads point to your work and, more importantly, what your follow-up

01:16:00.220 work will look like. So, again, I realize that to try to secure funding to do this from NIH

01:16:06.160 is slightly more complicated than trying to get bipartisan support on healthcare. But if we took

01:16:14.180 the economics out of the equation, if there was a $10 to $20 million pool that the National Institutes

01:16:20.200 of Aging said, you know, Dr. Caberlin, we want the definitive work on how to extend the life of dogs,

01:16:28.260 knowing that that's probably the best thing we're going to get towards humans, what do those

01:16:33.580 experiments look like? Right. So the study that we want to do is a five-year study with rapamycin.

01:16:40.800 And certainly, this is scalable, right? So what we have designed is a study with rapamycin.

01:16:47.640 There are other interventions coming down the pipeline that I think have the potential to be

01:16:52.620 as effective as rapamycin. So this doesn't necessarily, as long as you can do it safely

01:16:56.460 in pet dogs, you can test any intervention. And one of the things that I am now convinced is

01:17:02.160 owners are enthusiastic about participating in these kinds of studies. We've had more than 6,000 people

01:17:08.220 sign up through our website with no advertising at all to participate in the rapamycin trial.

01:17:12.840 In human clinical trials, which I'm more familiar with, recruiting cost is an enormous cost.

01:17:18.500 Yeah, it doesn't cost anything.

01:17:19.260 Yeah, you get free recruiting.

01:17:20.140 Right, right.

01:17:20.720 That's nice.

01:17:21.220 So having said that, the study we would like to do is a five-year study in dogs starting treatment at

01:17:26.980 middle age. So again, the dogs would come in, you know, at least six years old. We might push

01:17:31.980 that up to seven or eight years old and they'd, there would probably be a weight limiter like

01:17:37.140 I talked about before because again, big dogs age faster. So let's just say 40 pounds, six

01:17:41.940 years old, at least six years old. So there will be dogs anywhere from six to 10 or 11. Bring

01:17:47.540 in 450, 500 dogs or enough dogs to get 450 all the way through the study and look at not

01:17:57.340 just lifespan. But lifespan is actually a really important metric here, right? Lifespan is certainly

01:18:04.120 still the gold standard in the aging field. If we want to convince the scientific community

01:18:08.500 that this is affecting aging, it darn well better increase lifespan. It's also important

01:18:12.660 to owners for obvious reasons. And I think because as we talked about, euthanasia is really

01:18:17.340 what most dogs die from because the dog gets sick enough that the owner and the veterinarian

01:18:21.960 decide that it's time to put the dog down. I think lifespan is actually a really good metric

01:18:28.500 of healthspan in dogs because most of the time that's not an easy decision, right? Owners are

01:18:34.160 not going to put their dog down usually, especially owners who want to participate in a study like

01:18:38.480 this, unless that dog is really sick. So I actually think lifespan is really maybe more important in

01:18:43.280 dogs as an outcome measure than it is in mice. So lifespan is one of the key endpoints that we want to

01:18:49.580 look at. And then we want to look as broadly as we can at functional measures of aging that we

01:18:56.900 know are important in dogs. And so this goes way back to what we were talking about before. What do

01:19:02.540 dogs get sick with as they get older? So heart function, we will definitely look at in part

01:19:07.300 because that's what our preliminary data... Yeah, you showed that with a handful of dogs.

01:19:11.560 We have good evidence that we can detect and expect to detect improvements in cardiac function.

01:19:16.060 Activity. So one of the nice things about dogs is you can put a GPS tracker on their collar and get

01:19:22.140 very quantitative measures of activity. As the sensor technology improves, it might be feasible

01:19:28.080 to use a microchip instead of a GPS tracker on the collar to also get some physiological...

01:19:33.440 Or just throw a Fitbit on their res, you know.

01:19:35.320 Yeah, right. Well, I mean, that's really what the trackers on the collar are, right?

01:19:37.620 Yeah, yeah, yeah, of course.

01:19:37.980 But it would be nice to get some physiological measures in sort of continuous real time if we can.

01:19:43.260 So far, the sensor technology isn't quite there. At least that's my understanding. But we'll at

01:19:47.080 least get activity. And that'll give us a measure both of things like arthritis. So dogs that have

01:19:52.660 arthritis are going to be less active. And also muscle function. And also how well are they feeling?

01:19:58.700 Again, a dog is more likely to be active if it feels well. Now, obviously, that's confounded a

01:20:02.900 little bit by whether the owner takes the dog for a walk. But I still think total activity is an

01:20:07.000 important thing.

01:20:07.560 And is this a three-group study in your mind?

01:20:09.480 So the study that we're designing now is three groups. It doesn't have to be. But the three

01:20:13.840 groups are a placebo group, a short-term group. So six-month or year-long treatment. And then the

01:20:20.820 continuous treatment group. The reason for doing that, again, comes back to the mouse data in part

01:20:26.240 because we, as I mentioned, we've published that a short-term treatment in mice is enough to give

01:20:31.100 you large benefits on lifespan and at least some measures of health. And also because, again,

01:20:36.400 as we're thinking about ultimately bringing this to people, it's easier to envision a transient

01:20:44.160 or intermittent treatment than it is a continuous for the rest of your life sort of treatment.

01:20:50.240 So it would be single dosing?

01:20:52.000 Again, that's the way we have the study designed now. I can see a rationale for doing three months

01:20:57.060 on, three months off or some variation on that. We felt that the simplest thing to do first. So,

01:21:03.360 you know, it's always a balance between getting as much information as you can from a study like

01:21:08.040 this and doing things that are going to be where the complexity isn't so great that it outweighs.

01:21:12.960 Yeah, your power analysis could give you a study that's true.

01:21:15.840 So the design that we're working with now is one year on and then the rest of the time off. And so

01:21:21.640 those, so we'll get...

01:21:22.860 So you'll have one group that's five years on, one group that's one on, four off, and then placebo.

01:21:27.480 That's right.

01:21:28.040 Blinded across the board. 0.51

01:21:29.200 Yeah. It's all going to be a randomized double-blind trial, right? In addition to

01:21:33.700 lifespan and heart function and activity, we'll track cancer incidents.

01:21:39.880 Probably kidney function.

01:21:40.760 Kidney function. Get routine blood chemistry on the dogs probably every six months. We'll ask the

01:21:45.320 owners to bring their dogs in, get echocardiograms, collect serum for metabolomics and feces for

01:21:52.700 microbiome.

01:21:53.480 So what would that study that you just described cost if you could complete... So let's say you

01:21:58.580 had to screen 600 dogs to complete 450, which would actually be pretty good attrition.

01:22:02.840 So it would run about 5 million. That's about the budget for that study. Some of that depends on

01:22:09.440 what we can ultimately get the rapamycin for.

01:22:12.680 Because right now it's the street value is about...

01:22:15.040 The lowest we've been able to find is about $7 a milligram. Yeah. Now it may be possible if we

01:22:21.120 have a large study that we can identify a supplier that would cut us a deal for less. And the drugs

01:22:26.040 are actually a pretty large amount of the budget.

01:22:29.380 Yeah. The veterinary costs are the other large expenditure. Again, the nice thing about companion

01:22:34.200 dogs, it's unlike a study in mice in the lab, is they live with their owners. We don't pay cage

01:22:39.120 costs and things like that. The other major costs are going to be for...

01:22:44.100 The analysis.

01:22:44.640 The analysis. And also we need people to be able to communicate with the owners. Retention will be

01:22:52.420 important. I don't think it's going to be as hard as it is for some clinical trials, just based on

01:22:57.660 our experience. Even though it was a short-term trial, the owners that came into the short-term

01:23:02.000 trial, they were extremely...

01:23:03.980 This will be easier than communicating with patients in a trial, I suspect.

01:23:07.660 I think so. Yeah. And there's actually, I mean, it's kind of funny, but there's actually data that

01:23:10.900 dog owners are more likely to give their dog a prescription medication than they are to take

01:23:15.140 their own prescription medication. I would not doubt that for a second.

01:23:18.000 So yeah. So the owners that come into this study are highly motivated. And some of them are extremely

01:23:22.860 disappointed after the study ends when they find out their dog was in the placebo group. So,

01:23:27.080 but there is a communication component to this where we have to keep the owners engaged,

01:23:31.760 maintain communication. We'll be sending out regular surveys, but it needs to go beyond the

01:23:35.840 surveys. Now, would these all have to be dogs that live in the Northwest? They'd have to be able to come

01:23:40.360 in to... In fact, they almost certainly would not be. So the way that we're planning the study now

01:23:45.240 is that we will partner with five to seven veterinary schools around the United States.

01:23:50.940 Oh, that's great. And that actually, there's lots of reasons why that makes sense.

01:23:53.700 Well, diversity alone is great. Right. But veterinary cardiologists, there's not a huge

01:23:58.260 population of veterinary cardiologists out there. So if we were to try and do a study like this in the

01:24:02.680 Seattle area, I don't think that we have enough veterinary cardiologists to actually do the,

01:24:09.280 just the cardiology part of the study. So fortunately, you know, veterinarians are very

01:24:13.920 enthusiastic about participating in projects like this. And we have collaborators lined up at the vet

01:24:19.580 schools around the country. So it will probably be five or seven sites. Obviously, all of those sites

01:24:24.860 have to have veterinary cardiology, but almost any major veterinary school is going to have that.

01:24:30.900 And we prefer to work with veterinary schools that are in a suburban or urban area. Some veterinary

01:24:38.140 schools like our school in the state of Washington is all the way on the other side of the state from

01:24:42.640 Seattle. It's kind of out in the middle of nowhere. So that makes it harder to get owners to actually

01:24:47.300 bring their dogs there. Yeah. So our lead veterinarian is at Texas A&M Veterinary College. And so she 0.90

01:24:54.340 would be the head clinical person on the study. And that would probably be the site that the other

01:24:59.800 veterinary sites coordinate with. Yeah.

01:25:02.220 Well, Matt, I could sit here and have this discussion for another two hours.

01:25:06.080 It's fun stuff to talk about.

01:25:07.120 It is. And I really appreciate your time and your insights. And I think the work you're doing is

01:25:10.620 certainly what I would consider to be among the bodies of work that are at that tip of the spear,

01:25:15.820 because I guess I don't have a lot of hope we're going to get the answer to this question

01:25:19.780 directly. I think it's going to be an indirect triangulation of data. And I think to be able to do

01:25:27.600 this in companion dogs that live in our environment is going to be a really important thing. So let's

01:25:34.540 see if we can get that study done.

01:25:35.540 Yeah. I also want to, I mean, I think it's also important to at least note the impact of a study

01:25:41.260 like this would have on public perception, right? I mean, I think in the absence of any data,

01:25:46.620 and then the little bit of data we got from the phase one study, the amount of media attention that

01:25:51.580 we've gotten.

01:25:52.000 Yeah. I've heard you on NPR talking about this with Terry Gross.

01:25:54.940 Right. Has been huge. And so I tend to agree with you that it's going to be

01:25:59.220 challenging to generate enthusiasm for a double blind placebo controlled clinical trial of rapamycin

01:26:06.940 for healthy aging in people. Challenging is probably not even a strong enough word, but I think we do

01:26:12.960 need to, or we should at least not underestimate the potential impact if we're successful at accomplishing

01:26:18.960 this in people's pets, but that will have on public perception as well as perception among the broader

01:26:26.120 scientific community. I do feel like the field of aging research still has a bit of a reputation

01:26:33.280 problem among the broader scientific community. Part of that is historical. Part of that is because

01:26:39.540 there are some fringe elements that get a lot of attention, but that aren't scientifically credible.

01:26:44.500 I think that actually being able to show in dogs living in the human environment that we can modify

01:26:51.260 aging will have an impact not just on the public in terms of, I'm sure we'll get lots of media

01:26:56.980 attention, but also among scientists who might actually say, okay, aging research has arrived. I

01:27:03.340 think that's happening already. I think the TAME trial has actually been mostly a positive in that 1.00

01:27:08.020 sense, or the proposed TAME trial, I should say. But I think we still have some work to do.

01:27:11.860 The targeting aging with metformin. Yeah. Which is, I mean, I support that study. I think that it's

01:27:18.220 probably the right first study in this area because as you've already said, metformin, we know that

01:27:24.260 it's very safe, at least as far as a drug you might consider for a study like this. It's very safe.

01:27:30.080 There's very good human data suggesting that diabetics taking metformin not only have less

01:27:37.800 diabetes, but they have fewer other age-related diseases. So I think the human data is pretty

01:27:42.260 compelling. The downside to metformin, and I think one of the reasons why it has been a struggle to

01:27:47.340 get that study funded is that I think there's a perception that we already know about metformin.

01:27:52.300 It's not going to be completely surprising given the literature that's out there if it does have

01:27:56.860 relatively small effects on other age-related diseases. And then I also think it probably isn't

01:28:01.500 going to have that large of an effect. I could be wrong. I hope I'm wrong. But again, my view of

01:28:07.260 the literature that's out there is that metformin probably has modest effects, but they're not going

01:28:13.040 to be 20% increase in lifespan and rejuvenation of heart function and immune function. So I think

01:28:21.060 that that's probably the downside to that study. But again, because we're also battling this

01:28:26.440 perception of resistance to the potential for side effects when you're talking about treating

01:28:32.240 healthy elderly people, that's probably the right way to design this first study. The other thing

01:28:37.760 about the TAME trial, though, is it's not being done in healthy elderly people, right? The people 0.86

01:28:41.940 that they're enrolling have to have at least one age-related disease, and it can't be diabetes.

01:28:46.380 So even that is really not the gold standard study that we'd all like to see, whether it's metformin

01:28:52.200 or rapamycin or something else, we've got some work to do to get to where we can actually do that

01:28:57.360 study. And maybe the path forward is, as you said, individuals who are willing to kind of come

01:29:03.460 together and do these sort of self-experiments, if you can do it in a rigorous way where you're

01:29:10.820 actually measuring the things that are most important. There needs to be a model system that allows it

01:29:15.840 so that if you and I and Bob and Rick and John do it, we're, you know.

01:29:21.180 The downside to that kind of a model is that it's still going to be hard to convince just the

01:29:27.340 scientific community, the way that it is. It's much harder to convince them that it's real. Having said

01:29:32.380 that, if the effects are robust and you see it over and over and over in multiple people, we'll get

01:29:38.240 there eventually. So that may be the path that we end up taking.

01:29:41.220 You know, I'm not betting on that path. I'm betting on the path that you're describing as being

01:29:45.840 the shortest distance between these two points.

01:29:48.800 And I hope that as we start to do more of these studies in dogs and also the preclinical stuff

01:29:55.560 in mice, as we start to find functional measures that are improved over a relatively short timeframe,

01:30:03.280 that people will start to do some of these clinical trials that are feasible in people. I mean,

01:30:08.620 Novartis has already done it, right? You can do a clinical trial where you treat healthy elderly

01:30:14.980 people for six weeks, eight weeks, 10 weeks, and look at a functional outcome. And so if it's the

01:30:21.720 case that rapamycin rejuvenates immune function and rejuvenates cardiac function and delays or restores

01:30:28.560 alveolar bone levels in the mouth, right? Those are clinical endpoints that are impactful and could be

01:30:35.300 studied in a short clinical trial. So we may get a few of those that will kind of build this body

01:30:40.600 of evidence that rapamycin is having a similar effect in people. Again, the hard part is there's

01:30:45.620 not a lot of money in rapamycin. So I don't know who's going to fund those trials. That's the

01:30:49.600 challenge. So we need to identify. So either it's going to be rapamycin derivatives that are under

01:30:55.600 patent, like Novartis has developed and now Restore Bio is further developing, or it's going to be

01:31:02.060 alternative funding sources, whether that's foundations or wealthy individuals who recognize

01:31:07.780 the potential impact of this work and are willing to fund a clinical trial to actually start to look

01:31:13.920 at some of this. Well, Matt, thanks again. This was super interesting. Absolutely. And best of luck

01:31:17.700 with your continued work. Thank you. You can find all of this information and more at

01:31:23.320 peterattiamd.com forward slash podcast. There you'll find the show notes, readings, and links related to

01:31:29.440 this episode. You can also find my blog and the nerd safari at peterattiamd.com. What's a nerd

01:31:35.580 safari, you ask? Just click on the link at the top of the site to learn more. Maybe the simplest thing

01:31:40.300 to do is to sign up for my subjectively non-lame once a week email, where I'll update you on what

01:31:44.960 I've been up to, the most interesting papers I've read, and all things related to longevity,

01:31:49.680 science, performance, sleep, etc. On social, you can find me on Twitter, Instagram, and Facebook,

01:31:54.720 all with the ID peterattiamd.com. But usually Twitter is the best way to reach me to share

01:31:59.720 your questions and comments. Now for the obligatory disclaimer. This podcast is for general informational

01:32:04.780 purposes only and does not constitute the practice of medicine, nursing, or other professional health

01:32:09.640 care services, including the giving of medical advice. And note, no doctor-patient relationship is

01:32:15.240 formed. The use of this information and the materials linked to the podcast is at the user's own risk.

01:32:20.940 The content of this podcast is not intended to be a substitute for professional medical advice,

01:32:25.720 diagnoses, or treatment. Users should not disregard or delay in obtaining medical advice for any medical

01:32:31.140 condition they have, and should seek the assistance of their health care professionals for any such

01:32:35.700 conditions. Lastly, and perhaps most importantly, I take conflicts of interest very seriously. For all

01:32:41.600 of my disclosures, the companies I invest in and or advise, please visit peterattiamd.com forward slash about.

01:32:50.940 Thank you.

The Peter Attia Drive - August 20, 2018

#10 - Matt Kaeberlein, Ph.D.： rapamycin and dogs — man's best friends？ — living longer, healthier lives and turning back the clock on aging, and age-related diseases

Episode Stats

Summary

Transcript