In this episode, neurosurgeon and inventor, Kevin Tracy, joins me to talk about the importance of the vagus nerve and how we can harness its healing reflexes to improve our lives. This episode is sponsored by the Feinstein Institute for Medical Research.
00:11:28.220Yeah. That's the right answer. Wow. So now we can say, well, where are those signals traveling
00:11:34.860from the brain to the body and back? Some of them are in the vagus nerve and we can draw maps
00:11:39.300of those fibers from the, from the neurons in the brain to the fibers in the vagus nerve to the
00:11:45.940location in the intestine or the, or, or, or wherever else we're studying inflammation. But that's what we're
00:11:51.800able to do now. You can't, I, I, a computer model, even a very sophisticated AI model can't conjure that
00:11:58.940up. It has to start with the observations from these kinds of modern technologies.
00:12:05.000Uh, only because we're on the theme of computers, I'll ask one question about AI, and then we'll go back
00:12:09.920to how you chose that particular interest in, in, in studying that particular, you know, system. Uh,
00:12:16.740do you foresee ever a day where, uh, there is no need for Dr. Tracy performing the surgeries because
00:12:24.620the robots will be so, uh, their, their dexterity will be so, I mean, I saw recently on, I think it
00:12:31.060was an Instagram reel where these robots were, you know, were peeling like the, some unbelievably small
00:12:38.680minutia that you could have never imagined that they would have that level of dexterity. Is this
00:12:44.180something that is way, way in the future science fiction, or are many people in medical fields,
00:12:49.700whether it be in neurosurgery or certainly like in radiology, for example, where, you know, you have
00:12:54.440the pattern recognition that, that the AI can come up with that oftentimes beats the ability of the human.
00:13:00.960How does AI challenge or contribute or help, uh, neurosurgery?
00:13:07.460In the way that you just described, I think you're going to see, uh, who was it? Ernest Hemingway said
00:13:15.980change happens slowly and then all at once, something like that. I think that we're living
00:13:21.720through now the change happening slowly. And, and there's, there's a good reasons for change
00:13:26.980happening slowly in medicine. You know, even in your own personal case, you don't want your own
00:13:31.380doctor chasing every shiny new object every five minutes, right? You want, you want careful
00:13:36.360conservative thinking to be the, the first do no harm to that, to guide, guide the early adopters and
00:13:44.600the early, the early adopters and the early decision makers. That's all fine. Well, in that, in that
00:13:50.180scenario, change is already happening, right? As you said, in the, in the medical fields that are,
00:13:57.660that are highly image oriented, you mentioned radiology. There's also dermatology and pathology,
00:14:04.140even, even now, um, AI guided and computer assisted, if not computer, um, uh, I'm not gonna say
00:14:16.280independent, but computer deterministic evaluation of colonoscopies, right? There are products now
00:14:23.560out on the market where the, the, the camera is, is feeding all the information straight into
00:14:30.940a, um, a computer that's, that's reading every single image and picking up things, uh, polyps and
00:14:38.780potential tumors or other problems that the doctor may or may not have picked up. That means we're one
00:14:44.680step away from colonoscopies being done by swallowing the camera. Why do you need, why do you need the
00:14:49.480doctor to hold the tube if the, if the images are all being read after the fact by, by a computer?
00:14:55.480So these things are happening as we speak. I mean, if you, even in psychiatry, right, where you have,
00:15:01.840this goes back decades where psychiatrists have been scanning, uh, web searches, search engines
00:15:09.460to, to, to pick up diagnosis of early depression, anxiety, or even schizophrenia from, from kids as they,
00:15:18.420as they do their Google searches. And certain combination of, of words have been, are being
00:15:24.280studied as possibly being warning signs that this, this person should, should be helped, should seek
00:15:29.900help. These things are all happening and have already happened in some cases. So when you, when you
00:15:35.180look at the power of robot assisted surgery today, whether it's for endoscopic surgery with GI surgery
00:15:42.860or, or, or urology, uh, procedures in, uh, in, in another gynecological procedures or neurosurgery,
00:15:50.280as you said, I think absolutely robots are going to be, are going to be more and more capable of doing
00:15:56.560more and more of the, of the surgical intervention with less and less of the human intervention.
00:16:01.780Yeah. I'm, I'm just brainstorming here. And so I'm thinking back of some of the research that I had
00:16:07.260studied when I was a doctoral student, where if you show people that the likelihood of you being
00:16:14.020accepted into a program, if you use an actuarial model, just a mathematical model actually results
00:16:21.740in better outcomes than if you use a human decision maker, because the human decision maker,
00:16:26.840maybe their blood sugar is low because they're hungry and now they're pissed off. So the exact same
00:16:31.720file that I might've said, yes, okay, get him into Princeton. Now I say, screw this guy. He's not going to get
00:16:36.720into Princeton. So even when you tell people, Kevin, that it would be in their best interest to use an
00:16:43.560actuarial model, they don't feel comfortable in giving their future away to a cold mathematical model.
00:16:53.560Might it also be the case that even if we ran studies showing that the likelihood of a neurosurgical
00:17:00.800robot of making an error is lesser than a human one, do you think that it would be the case that
00:17:08.060people would have reticence? I want a real human being working on my brain, not some cold algorithm.
00:17:15.500Does that seem reasonable for me to expect that they would have that bias?
00:17:18.960I'm not a psychologist. I'm not an expert in this space, but I will tell you first,
00:17:26.880I saw a statistic recently. And like I said, I don't know if it's true, but it was from a reputable,
00:17:32.440it was from a reputable journalistic source that a significant percentage of job applications
00:17:38.940to major corporations today are already being screened by AI. So in one way, you can say there's
00:17:46.080already been an abdication to that model, the actuarial model. And I think the same thing
00:17:51.020happens in colleges, where a large number of the college applications are already being screened
00:17:57.640by AI. And as far as what people will prefer when they seek their healthcare, I mean, there's the old
00:18:06.560neurosurgery joke, right? If you ask a neurosurgeon who the three best neurosurgeons in the world are to
00:18:12.580solve that patient's particular problem, they have a very difficult time thinking of the names of the
00:18:17.520other two. So it's complicated. You know, I think medicine is, I think as doctors, and I no longer
00:18:29.880treat patients, I'm not practicing now. I retired from my clinical practice. I run my research lab now
00:18:35.800full time. But when you think about what we're going through, and maybe this is an important
00:18:41.740punctuation point on this whole conversation. AI is here, AI is not going away, AI is going to be
00:18:48.000more and more involved in all aspects of our lives. That's, and it will replace some, the way we do some
00:18:55.440things. Our obligation now as healthcare leaders, and providers, is to make sure that we insert the
00:19:04.300humanity, the humanity, the humanness, the human touch into healthcare, because AI can't do that. And it needs to
00:19:15.380be sort of constantly reiterated, discussed, and, and reminded that patients with the highest trust in their, in their
00:19:27.600physicians, or in their healthcare providers, have better outcomes. This has been studied, improved 10 ways to Sunday. If you have a high, those
00:19:37.880patients with the highest confidence, the most belief in their caregivers have better outcomes. There's something about the human
00:19:44.820touch and the interaction of compassion and caring, especially when, frankly, medicine runs out of options, you still have that
00:19:56.000compassion and caring, that we cannot lose sight of it has to remain on the critical path of what we do in all of medicine and
00:20:04.420Beautiful. I'll come back to your book, and one particular term, which I'm sure you'd be happy to talk about, bioelectronic medicine. So let's, let's
00:20:13.020table that for a second. But since we're talking about neurosurgery, do you mind if I ask why you decided to, to leave your
00:20:20.140clinical practice? Because I was got one of the questions I was going to ask you is that you wear many hats, which one do you
00:20:26.080prefer? And why? And has that preference shifted across your career? And it would seem to me that maybe now, perhaps you're more
00:20:34.100excited about your research than you are about the clinical. So walk us through all those decisions. And again, I ask these
00:20:39.280questions, first, because I'm interested in people that I talked to. But as a psychologist who studied psychology of
00:20:44.960decision making, these types of decisions always fascinate me.
00:20:49.200Well, I mean, it's a it's a long story. But but to tell it in brief, my mother died of a brain tumor when I was five. And it came as a
00:21:00.340very sudden surprise to all of us. My my father, her father was a PD, a professor of pediatrics at Yale and New
00:21:09.120Haven, Connecticut, came as a surprise to him, my, my infant baby sister and my younger brother, all of us were
00:21:18.040obviously devastated by this. And so all my life, I've been very sensitized to what is neurosurgery and what is the
00:21:25.400impact it can have. By the time I was a teenager, I had really decided I wanted to be a scientist.
00:21:31.780All my life, I knew pretty much as long as I can remember, I wanted to have a laboratory, I wanted to
00:21:37.400invent things. And I wanted to understand how to invent things that could lead to cures. I just thought it was
00:21:44.040such a fascinating path. And so I was on that path as a chemistry major in college at Boston College, and then
00:21:52.900making a career decision of getting a PhD in chemistry, where I could then go learn to make
00:21:58.140pharmaceutical drugs, or getting an MD, and learning how the diseases actually work. And that's the path
00:22:07.300I chose, obviously. And then in medical school, you have to pick surgical versus non surgical
00:22:11.640specialties. The decision to become a neurosurgeon, you know, it's certainly an interesting story now
00:22:18.440looking back, but I didn't become a neurosurgeon, because my mother died of a brain tumor when I was
00:22:24.200five, I became a neurosurgeon, because when I finished my training in the early 1990s, so as a
00:22:33.500fully trained neurosurgeon, the neurosurgery departments had a lot of resources. And this was a
00:22:41.340time when the departments, whether it was neurosurgery, or medicine, or general surgery, the departments
00:22:46.260with the most resources had the biggest opportunities to support research laboratories.
00:22:51.540And so I could combine a career doing neurosurgery one day a week, and then running a big laboratory
00:22:58.020for the other, you know, six days a week, which is what we all did. We worked, we worked in the clinic
00:23:04.420a day a week, and then we became obsessed with our laboratory projects. So I always had a career that
00:23:11.280combined these interests of doing research to discover ways of trying to help people and
00:23:17.280understanding how diseases worked. And in neurosurgery, I wouldn't, I wouldn't trade a minute of it away. I
00:23:23.920loved learning how to do neurosurgery. I loved interacting with the patients. I love doing
00:23:28.400neurosurgery. But I wouldn't trade away a minute of, of the years I spent in the laboratory working with
00:23:34.900brilliant graduate students and brilliant postdocs trying to unravel how, how, how, how the universe
00:23:44.580is assembled, you know, take it apart piece by piece, and then build it back together again,
00:23:49.700so that you can, you can invent new therapies. And we've been fortunate because we've, we've had a
00:23:54.900couple of, of, of wins in that, in those stories. And, and it wouldn't have happened, I think, if I wasn't
00:24:00.500acutely aware from my training of, of, of how, how disease affects people, and, and how the importance
00:24:10.500of the brain and the nervous system in those connections and those responses, but it also
00:24:14.980wouldn't have happened if I hadn't taken the time to spend years in the laboratory and learn the basic
00:24:19.140science principles of how to, how do you do an experiment to take apart a complex molecular
00:24:24.660interaction and break it down to the simplest, the simplest assemblage, and then reconfigure
00:24:31.460it in a way so what's broken is now fixed. Beautiful. When I was deciding after my undergrad
00:24:37.860what to, to go into, one of the things that I thought of doing was to do an MD, PhD, I was very
00:24:45.460much interested in potentially psychiatry, just because I was interested in behavioral sciences and so on.
00:24:51.620And the reason why I would have been interested in doing an MD, PhD is exactly what you intimated,
00:24:56.020which is, I would have liked to, once in a while, you know, be doing the helping others. But I was,
00:25:02.260at heart, I was somebody who wanted to solve puzzles. And so I thought that the MD, PhD would work out.
00:25:08.180Later, I ended up not pursuing that goal, because I was quite disheartened, and you'll, you'll tell me if,
00:25:13.620if I'm right or wrong, by the amount of what seemed to be quackery in the mental health professions in
00:25:22.420general, right? So it went from all the Freudian stuff, where it was all environmental, mommy didn't
00:25:28.340hug you enough, or maybe she hugged you too much, that's why you're schizophrenic, to completely the
00:25:32.900opposite, where everything became pharmacological. And like most things in life, the sweet spot is
00:25:38.340somewhere in the middle, a bit of listening to the patient, and maybe being empathetic.
00:25:43.620And when need be, giving that pill is probably the right. So I could have followed you in that MD,
00:25:49.140PhD, I know you didn't, you didn't pursue a PhD, right? You got you got your right. All right,
00:25:54.420let's go back, go back to the book. And let me mention it again. I don't have the physical copy.
00:25:59.060And I think your your assistant sent it to me, to my university address here, where I'm on leave from
00:26:05.060my university. But let me read it, the great nerve, the new science of the vagus nerve and how to
00:26:10.660harness its healing reflexes. Tell us all about bioelectronic medicine.
00:26:17.300The book tells the story of something I mentioned briefly earlier, which is that signals in the vagus
00:26:23.780nerve can turn off inflammation. And we made this discovery quite by accident, you asked about the
00:26:30.980origin story in my own career case, the origin story of this accident is that we were studying
00:26:38.100animals that had a stroke, cerebral infarction. And the idea was, the stroke causes inflammation
00:26:46.100in the brain, which it does do. And we put a molecule into the brains of these animals to turn
00:26:51.460off the inflammation, which it did do. And we were very happy with that. And that was what we expected
00:26:57.540to see. What we didn't expect to see was when we put these molecules in the brains of these mice,
00:27:05.140this is back in the 1990s, that the inflammation in the body of the animals was also turned off.
00:27:11.780So this was a real head, a real head scratcher. This is a real WTF moment. Because there was no way
00:27:18.420to understand how a molecule in the brain of a mouse would turn off inflammation in the body of the mouse.
00:27:24.020And the key experiment came when we cut the vagus nerves that were carrying the signal from the
00:27:31.620brain to the immune system in the body. And that was a real sort of, as I said in the book,
00:27:39.460we don't say Eureka in the lab anymore. We say, holy shit. Because once we realized there were signals
00:27:46.340in the vagus nerve that were acting like the brakes on your car to stop inflammation, then we realized we
00:27:53.860should be able to make a nerve stimulating device, a small battery powered electrode, and implant it in
00:28:02.100the vagus nerve of humans, because we already knew that could be safely done. As a neurosurgeon,
00:28:08.100it was being done to treat epilepsy patients back in the 1990s, so it's safe. And if we could do in humans
00:28:16.020what we were doing in the mice, then we should be able to treat conditions like rheumatoid arthritis and
00:28:21.460inflammatory bowel disease. So back in 2007, I started a company called Setpoint Medical with a colleague from
00:28:28.020Harvard named Shaw Warren. And now 18 years later, literally just a couple weeks ago here at my center, we
00:28:38.020implanted two patients, because the FDA approved this idea as a therapy for rheumatoid arthritis around the
00:28:47.940first of August. That whole story made the front page of the New York Times about three weeks ago.
00:28:54.500Wow. And the reason is that this now is a therapy that will be available to people with a very
00:29:03.460serious condition, rheumatoid arthritis, that's currently treated with very dangerous, expensive
00:29:08.660drugs called biologics, things with names like anti-TNF and anti-IL-1. And these drugs, right now,
00:29:16.420we're trading off the treatment of inflammation with these drugs with the side effects of immunosuppression.
00:29:22.420And with the side effects of potentially going bankrupt, some people have to spend $50,000 to
00:29:28.660$100,000 a year on these drugs in the United States. And they only work half the time. So now the question
00:29:35.140is, will some of these patients who are not gaining benefit from these expensive, dangerous drugs, will
00:29:41.060some of them have a immunoregulator, it's called, implanted in their neck as their primary therapy for
00:29:48.980rheumatoid arthritis. And the early data are very, very promising. That's what's happening as we speak.