EP. 131: AT THE EDGE OF PRECISION MEDICINE

WITH EUAN ASHLEY

Sep 26

WITH EUAN ASHLEY, MBCHB, DPHIL

A cardiologist, Chair of Medicine at Stanford University Medical Center, and a pioneer in precision medicine shares his vision for a future of medicine fueled by powerful genomic data and artificial intelligence, while grounded in the human touch.

Listen Now

Episode Summary

Precision medicine — the approach to health care that involves tailoring medical interventions to an individual's genetic makeup, environment and lifestyle — promises to deliver the right treatment to the right person at the right time. From preventing diseases decades before they appear, to specially designed cocktails of cancer drugs, to genetic modification of rare diseases, many of these applications sound straight out of science fiction.

At the forefront of precision medicine and medical genomics is Euan Ashley, MBChB, DPhill, Chair of Medicine at Stanford University Medical Center. A cardiologist and intensive care physician by training, Dr. Ashley has pioneered the use of genetic sequencing to identify risk factors for heart disease and new treatments for rare diseases. He is also the author of The Genome Odyssey: Medical Mysteries and the Incredible Quest to Solve Them (2021).

Over the course of our conversation, we discuss his path from growing up in a small Scottish town to now working at the cutting edge of medicine, the excitement and fulfillment he experiences as a clinician in the cardiac intensive care unit, remarkable patient stories of healing and resilience, the future of precision medicine, why he is optimistic about the development of artificial intelligence, and more.

Born in Scotland, Euan Ashley, MBChB, DPhil graduated with 1st class Honors in Physiology and Medicine from the University of Glasgow. He completed medical residency and a PhD at the University of Oxford before moving to Stanford University where he trained in cardiology, joining the faculty in 2006. His group is focused on the science of precision medicine. He is best known for his work helping establish the field of medical genomics. His team developed some of the earliest tools for the interpretation of the human genome in the context of human health.
He founded the Clinical Genomics Program and the Center for Inherited Cardiovascular Disease at Stanford. He was the first co-chair of the steering committee of the national Undiagnosed Diseases Network. He was a recipient of the National Innovation Award from the American Heart Association and the NIH Director’s New Innovator Award. He was recognized by the Obama White House for his contributions to Personalized Medicine. In 2018, he was awarded the American Heart Association Medal of Honor for Genomic and Precision Medicine. He was appointed Stanford Associate Dean in 2019 and became the inaugural holder of the Roger and Joelle Burnell Chair in Genomics and Precision Health in 2021. In 2023, he became a Fellow of the John Simon Guggenheim Memorial Foundation. He is co-founder of several companies including Personalis, Deepcell, and Svexa. His first book The Genome Odyssey - Medical Mysteries and the Incredible Quest to Solve Them was released in 2021. Father to three Americans, in his spare time, he pilots planes, tries to understand American football, plays jazz saxophone, and conducts research on the health benefits of single malt Scotch whisky.
In this episode, you will hear about:
• 2:24 - Dr. Ashley’s path to medicine and to cardiology
• 7:19 - What life is like working in the CCU
• 21:34 - How the Undiagnosed Diseases Network was founded and what it does
• 33:22 - An overview of precision medicine
• 38:09 - The impact that genetic testing and genomic medicine is having on modern medicine and where it could go from here
• 45:00 - Dr. Ashley’s thoughts on how AI will change the field of medicine
• 51:40 - Making access to medical advancements in AI and genomics more equitable
• 1:04:39 - Dr. Ashley’s advice for healthcare professionals in training
Henry Bair: [00:00:01] Hi, I'm Henry Bair.
Tyler Johnson: [00:00:02] And I'm Tyler Johnson.
Henry Bair: [00:00:04] And you're listening to The Doctor's Art, a podcast that explores meaning in medicine. Throughout our medical training and career, we have pondered what makes medicine meaningful. Can a stronger understanding of this meaning create better doctors? How can we build healthcare institutions that nurture the doctor patient connection? What can we learn about the human condition from accompanying our patients in times of suffering?
Tyler Johnson: [00:00:27] In seeking answers to these questions, we meet with deep thinkers working across healthcare, from doctors and nurses to patients and healthcare executives those who have collected a career's worth of hard earned wisdom probing the moral heart that beats at the core of medicine. We will hear stories that are by turns heartbreaking, amusing, inspiring, challenging, and enlightening. We welcome anyone curious about why doctors do what they do. Join us as we think out loud about what illness and healing can teach us about some of life's biggest questions.
Henry Bair: [00:01:02] Precision medicine, or the approach to health care that involves tailoring medical interventions to an individual's genetic makeup, environment and lifestyle promises to deliver the right treatment to the right person at the right time. From preventing diseases decades before they appear to specially designed cocktails of cancer drugs to genetic modification of rare diseases, many of these applications sound straight out of science fiction. At the forefront of precision medicine and medical genomics is Doctor Euan Ashley, chair of medicine at Stanford University Medical Center, a cardiologist and intensive care physician by training. Doctor Ashley has pioneered the use of genetic sequencing to identify risk factors for heart disease and new treatments for rare diseases. He is also the author of the 2021 book The Genome Odyssey Medical Mysteries and the Incredible Quest to Solve Them. Over the course of our conversation, we discuss his path from growing up in a small Scottish town to now working at the bleeding edge of medicine, the excitement and fulfillment he experiences as a clinician in the cardiac intensive Care unit. Remarkable patient stories of healing and resilience. The future of precision medicine, why he is so optimistic about the development of artificial intelligence, and more. Euan, thank you for taking the time to join us and welcome to the show.
Dr. Euan Ashley: [00:02:30] Yeah, thanks for having me. I'm really delighted to be here.
Tyler Johnson: [00:02:33] I think he's only the second person we have had on the program who was previously my attending, back when I was a trainee in the internal medicine residency program here and now. So we usually have our guests start off by telling us, how did you end up becoming a doctor? What led you to entering the field of medicine.
Dr. Euan Ashley: [00:02:58] Well, wanting to be a doctor is really one of my first memories, really, in life. I certainly recall being about 5 or 6 years old, and friends of my parents asking me what I wanted to be, and I would tell them I wanted to be a doctor. Now, why is a little less clear at that age? But as I grew up, I think it certainly became clear that somehow, and I think many of us feel this, don't we? We were just sort of somehow put on this earth to try and help other people, and some element of that was part of it. At school, I was always the guy who would, you know, be called over when somebody grazed their knees or was bleeding on the playground. And, you know, I then learned first aid as an early time. My. I mean, I think that my parents were a huge influence on that. My dad was a GP in our local town for really his whole career and still lives there, although he's been retired for about 20 years now. And my mom was a midwife, not with us anymore, but she was a midwife also in actually the neighboring town. And so growing up, I was, I think, very influenced by both of them. I actually would go out on as a kid in house calls, not we don't do too many house calls anymore, but I would do house calls with both of them at various times. And so I was even in patients houses as a kid and seeing medicine sort of being performed in that setting. And it was really inspiring to me. And so I never really considered I, in fact, I never considered doing anything else. It was all I ever wanted to do.
Tyler Johnson: [00:04:21] Wow. Then what about, you know, we've had a handful of cardiologists on the show, although I think that you are the only at least as far as I'm aware, you're the only doctor that I know of that we have had who is a cardiologist who's also an intensivist or takes care of patients in the cardiac critical care unit. Tell us a little bit about how you ended up in cardiology, and specifically how you ended up being a doctor who takes care of critically ill patients with primarily cardiologic problems.
Henry Bair: [00:04:50] Also, because practicing in the cardiac ICU is so different from making house calls in a small town.
Dr. Euan Ashley: [00:04:58] Well, it's interesting there are some parallels between the doctoring I do now in the outpatient clinic, actually, and the work that I saw my dad do. I mean, I saw him look after generations of families where he would quite literally deliver a baby, and that baby would grow up and have another baby, and he'd deliver the next one. And so that was a pretty unusual and I think as I moved into subspecialty care and I'll touch on why why cardiology in a moment, I kind of felt that that was something I would have to lose. But interestingly, we have this center for inherited cardiovascular disease at Stanford. So we really do genetic medicine for inherited heart problems. And one of the absolute joys of that practice is that we get to meet whole families and including generations of families. And now I've been doing this long enough that we I have had the chance to see some little kids that I knew as as babies of some of our patients start to grow up, certainly into teenagers, and go off to college. So that's an unusual joy for me. But the heart to get to that, I remember distinctly, kind of the moment that I realized I had to become a cardiologist, and it was in the physiology class. So pre-med is this is in Glasgow in Scotland, so very traditional medical school education with preclinical and then clinical.
Dr. Euan Ashley: [00:06:15] And this was preclinical physiology fascinated me already because I love function and just understanding how things work. And it was the practical class where we got to see and tend to a heart. And so they literally took out a beating heart from an animal. Thankfully, we didn't get to see it, but it was hung on a little cannula and we had to look after this heart for the next 4 to 6 hours and keep it alive. And I remember just staring at this beautiful organ, this symphony of of kind of biology in front of me, its synchronization, the fact that the heart makes sounds and the fact that we could make it go faster with adrenaline or slow it down with beta blockers. It just absolutely entranced me. And from that moment on, I just knew I had to try to learn every single possible thing about this organ, and that has what has driven me. I mean, it didn't hurt the fact that Scotland has a very high rate of heart disease, as does most of the world at this point, and it's the biggest killer in society. And working on big problems is always also motivating. And that I had members of my own family with heart disease. But the actual moment was staring at that Langendorff heart which stayed with me all these years.
Tyler Johnson: [00:07:28] You know, I have to say two things. The first one is that I remember so at at Penn, where I went to medical school, we did about four months of sort of pre preclinical where we talked about genetics and general principles of physiology and whatever. And then we had a year of what we called organ blocks, where we looked at each of the, you know, the major organ systems in the body. And the first of the major organ blocks was cardiology, and the thing that I remember to be so bewitching about it was that the whole thing just made sense, right? Like when you got done, like later, six months later or something, when we did what we called brain and behavior, which was our neurology and psychiatry block, I remember we got done, I crammed for the final, I took the final, I did fine, but I remember getting done and thinking, I have no idea what what goes on in the brain or how any of those things work or any of what I just learned, even though I apparently learned it well enough to do fine on the exam. I have no idea what is going on between my ears. But in contrast, when I learned about the heart, just like you were saying, like it was so fascinating to be able to like, look at this EKG versus this EKG and tell, like, where were the atria contracting and where were the ventricles contracting, and why was the distance between it shorter or longer, or how could you make it faster or slower? What did it mean when the street wave elevated or whatever the thing was? And then the same thing looking, you know, learning V equals IR and cardiac loop physiology and all of that stuff. It made you sort of feel like a master of human physiology. Right. Because you could give inotropes or vasodilators or whatever the thing was and sort of be like, if I give this drug that has this effect, it will have this other effect on how the patient's doing. I thought it was remarkable.
Tyler Johnson: [00:09:11] And then by the same token, I remember when I was an internal medicine resident, kind of like you, I've wanted to be an oncologist for pretty much as long as I've wanted to be a doctor. But I had my my last rotation of my intern year and my second rotation of my second year as a resident were both a month in the CCU. And this is back when we were doing, you know, Q4 30 hour call. I mean, it was a big deal, intense experience. But I loved it. Like I was, you know, a hair's breadth away from switching to cardiology because the experience of working in the CCU was so remarkable. And so, you know, we have some people who are not even involved in medicine and then people at all levels of training, premeds, medical students, whatever, who listen to the program. Can you just paint a little bit of a picture for people who may not be familiar? What are the kinds of patients that you're generally taking care of in the CCU? And what are some of the remarkable things that you do and the remarkable recoveries that you see in the CCU pretty frequently?
Dr. Euan Ashley: [00:10:15] Yeah, although it was the organ itself and the physiology that you just very nicely described that drew me in to the cardiovascular system, what keeps me there and keeps me going back every, every year and every few months is the the critical care unit. And it's the patients that we get to interact with there. And so of course CCU grew up around the time they weren't always CCU. They grew up really around the time when we first realized that patients were having heart attacks, myocardial infarctions and dying of arrhythmias afterwards. And it wasn't that we had great tools for treating the arrhythmias, We had some pretty basic pharmacology, but grew up to look after that patient population. Of course, in a world where we can rapidly take patients for primary angioplasty, for essentially to place a stent and clear up the artery that is blocked and causing the heart attack. Often patients not always, of course. Plenty of complications, but often those patients can come to the CCU and they're there for a very short time and we move them back out. Some of them even are looking to leave the hospital the next day. So the original reason for for having a CCU has changed a little bit. But of course the range of patients has grown as we have found ourselves able to do other things traditionally so that that group of patients who had had a heart attack would be one patients who either for that reason or for other reasons, like having a genetic heart condition, who were having dangerous rhythms of the heart or another, especially if they're intractable, and maybe they come from another hospital where they've already done the standard things to try to reduce those rhythms.
Dr. Euan Ashley: [00:11:46] The patient's still suffering, and maybe every time they go into the rhythm, they either feel terrible or they literally lose their pulse. That's another group of patients who come in. The heart is a is a pump. It's much more than that, as we've already described. But it certainly is a pump first and foremost. And so the patients who have an inability now for that pump to push enough blood around their body who either have heart failure, as we refer to that, or cardiogenic shock at the most extreme end, where their organs are starting to fail because their their heart is not pushing out enough blood. That is now a very significant group, maybe even the major single group that we see. And there's a variety of others, of course, just because you have, let's say, a heart condition doesn't mean you're not at risk of an infection. And so when we have combinations of conditions, for example, patients with sepsis and infection in the bloodstream and heart failure, they do particularly badly and feel particularly ill. And so we have quite a few patients in that category.
Dr. Euan Ashley: [00:12:43] What's changed, of course, over the years, as well, since you were there at the beginning, is an ability for us to keep patients alive in ways that we otherwise wouldn't have done. And that has been, you know, really dramatic change in the past. If a patient's heart stopped in the CCU, we would do more or less the same things that you would do if their heart stopped and they collapsed in front of you at a at a mall, which is to say cardiopulmonary resuscitation, CPR. And of course, we'd have a range of tests. We could send, blood tests we could send and other drugs that we could give pretty much immediately. But of course, if emergency services come, those sorts of things would be available. The big difference actually, is really mechanical support, because these days, if somebody with a heart problem has a cardiac arrest, then we make the regular call to the cardiac arrest group who come running. Usually if you if you see, of course on TV and in real life, they actually do run one of the few times anyone is running in the hospital. The cardiac arrest team runs to the bedside, but now we have mechanical support which can take over the pumping of the blood. So one of the systems is called ECMO. Ecmo for extracorporeal membrane oxygenation. And it's basically a heart lung machine. It literally takes over the the function of the heart and lungs.
Dr. Euan Ashley: [00:13:58] And if you've been in a room, as we have many times, and had that transition happen where and it's not, I mean, the TV doesn't regularly reflect the reality of what we do, but there's some amount of the drama of a cardiac arrest that is accurately reflected in the TV shows is a lot of noise. It's a lot of people. Obviously, there's CPR happening in the middle, the patient at the center of it all, you know, a whole bunch of people in the room. And to go from that position to the one where suddenly a heart lung machine has taken over the function of the heart and lungs of the patient, and suddenly the volume level drops. The room becomes quiet, calm, takes over, is pretty remarkable. And that now is a significant component of many cardiac arrest protocols. We get the patient and we the phrase we use is that we crash them onto ECMO. And crash is probably the right word, actually, because it happens pretty fast and with a lot of noise. But the moment the machine is turned on and everyone steps back, there's no longer any need for CPR. The noise level goes down and things are much more calm. So that's a little bit of some of the changes that have happened over time. And some of the patients that we see in the in the CCU.
Henry Bair: [00:15:06] Yeah. So I actually worked in the CCU earlier this year as part of my internal medicine intern year. I was there for five, six weeks or so. Being a person on the road to ophthalmology, which is what I'm doing now. I went into the CCU honestly kind of dreading it. I expected it would be Byzantine, abstruse, totally outside the realm of what I knew and what I'd need to know for the rest of my career. Because let's face it, placing a central line or resuscitating a failing heart isn't quite called for in the day in, day out of ophthalmology. But to what Tyler was saying, I found that I appreciated how I could understand mechanistically what was going on. I remember learning about ECMO, as you described, but also balloon pumps inserted directly into the aorta of patients that inflate and deflate in sync with the beating heart in order to support blood flow and devices placed directly inside the heart to move blood along. Then there are all the medications. You can put people on continuous meds to keep blood pressure up or down, to stabilize the heart, to remove extra fluid building up in the body, and on and on. And I recall thinking not necessarily how easy it was, but how intuitive it was to understand everything that was happening, to have all those dials and settings right in front of you. And you can adjust the dosing and the flow rate and almost instantaneously see how the patient's physiology was responding. It was fascinating. On the other hand, it also struck me that unlike most other places in the hospital, the CCU was a setting in which if we stopped what we were doing to the patients, they'd be dead on the order of minutes to hours because of how intensively we were supporting their hearts. It's honestly quite humbling and breathtaking.
Dr. Euan Ashley: [00:16:54] I'm so glad to hear that. And I mean, it is an incredibly rewarding arena in which to practice. I mean, you have all the elements of intensive care medicine, the challenge of making sure you're communicating well with a patient who may be sedated or a family member. But if you do things right and things go well, that patient will wake up and they'll actually stay in your care. One of the things about intensive care that is a little unfortunate, in a way, is that often when they leave the intensive care unit, they go somewhere else and into the care of another doctor. One of the things I've loved about being an intensivist in the cardiac critical care unit is when the patient wakes up, you're still their doctor, and then to the point where they actually leave the hospital, they can become your outpatient, which is just something that I really have enjoyed over the years. And of course, not every story ends that way. But when it does, I have patients that I saw and first met when they were in extremis, really in a situation where there was single digit percentage chance of them surviving by any reasonable metric. And I am seeing them in clinic living a life, you know, ten years later, 15 years later, telling me about their grandchildren, showing me pictures of the vacation they had. And there are very few things in life more rewarding than remembering that night when you were up doing your best to keep them alive and then seeing pictures of their grandchildren, you know, like a decade later.
Tyler Johnson: [00:18:12] Well, and I think to that point, too, the thing that really struck me, and this happens in the. So for those who are still, you know, making their way into medical training or maybe are not going to have medical training in many hospitals, there's a differentiation between the medical ICU, the surgical ICU and the cardiac ICU. They might have slightly different names, but there are sort of different places. And oftentimes in the medical ICU, patients present with complications of underlying chronic illnesses that don't have great solutions to the primary root problem. Right. So if someone has widely metastatic cancer, for example, unless they're a newly diagnosed person who hasn't gotten therapy yet or something, it can often be the case that those underlying problems are intractable. And sometimes, of course, that's also true for patients who end up in the cardiac critical care unit. But there are also patients who either get ventricular assist devices or get a heart transplant, or have some other sort of major intervention where there sometimes are fixes for the root problem. And so you can have patients who literally, you mentioned earlier, patients who come in in heart failure or who are in cardiogenic shock. So either they are quite literally drowning in their own blood because the heart can't pump the blood out of the lungs in effect, or they are in shock because the heart isn't strong enough to pump the blood to perfuse the rest of the organs, and then if they get the appropriate kind of ventricular assist device, or especially if they get a heart transplant, then they can come in a year later, and now they're out hiking in the hills and living a relatively normal life, because that root problem actually has been addressed. And that, I mean, that kind of a Lazarus like almost rising from the dead is genuinely remarkable to see.
Dr. Euan Ashley: [00:20:01] Yeah, it is absolutely remarkable. And, you know, certain patient stories stick in your mind. One from the early days of my practice was we'd seen this tall African American woman who had sarcoid disease affecting both her heart and her lungs. And she'd been an athlete, like a college setting high school athlete, and then had gotten to the point because of this disease where she couldn't walk across the room. I remember actually, just before we put her on the transplant list, she could barely get out of the chair to walk across the consulting room. And I remember really distinctly as I think she went through and got a transplant, did very well, got back out of the hospital. It was not much longer. Maybe it was four weeks after her transplant and it so happened. It's not always the case, but it so happened. The post-transplant clinic, which is not the one I do in this case, is a heart lung clinic, was on the same day as my regular clinic. And so she was there in the clinic on the same day, and I saw her from the other end of the clinic, and she sprinted like she was like a gazelle, like, you know, the speed at which she, she basically sprinted down the clinic corridor and gave, gave me and my, my team a big hug. That's, you know, just one example. But it's is remarkable, this, this gift of life, that organ donation is. And now of course, if we could fix the disease upstream, that would be even better. But you're right. People can be literally close to death and unable to really live and then go to the point where they're climbing Kilimanjaro. You know, they're sending us photos from their hiking expedition in the Andes. So it really is, you know, a remarkable thing. And just another part of the sort of continuity of care and looking after those patients, that's really rewarding.
Tyler Johnson: [00:21:43] So speaking of getting to the root of diseases, I know that sometimes when I'm taking care of patients and they're really frustrated that the medical system is not doing a better job at helping them to feel better. Occasionally I'll draw sort of a random, you know, sort of a block on a piece of paper. And I'll say, let's pretend that this is all of the bad things that can happen to a human body, and then I'll draw a relatively small circle inside of that block and say, this is the area within which we can fit all of the things that we can reasonably diagnose. And then I'll draw an even smaller circle inside of the bigger circle and say, and this is the area that encompasses all of the things that we can actually helpfully treat, right? Which is to say that there are a lot of things that people are clearly suffering, and we can't even figure out what's going wrong, let alone how to fix it. Yeah. And so I want to pivot now. And one of the things that you have done in addition or as part of however you want to think about it, being a cardiologist is that you have and you mentioned this briefly earlier, gone very deep into what you just called genetic medicine. And partly you have done that through inherited cardiomyopathies, which we can talk about that as well. But specifically you, I think, helped to found a network for people with undiagnosed and theretofore undiagnosable diseases, where you developed a framework for trying to look for patients who might be helped by trying to find specific genetic causes of what ailed them. So can you talk a little bit about the genesis of creating that network? Sort of what the idea behind is it is and sort of how how that part of what you do, how that works.
Dr. Euan Ashley: [00:23:30] Yeah, absolutely. And the Undiagnosed Diseases Network is, at its core, a group of physicians and scientists collectively engaged in trying to solve some of the hardest medical mysteries around, some of the hardest cases, and the patients who have gone from doctor to doctor to doctor looking for answers where they don't have a diagnosis, they don't have a label, they don't have anyone have an understanding of their symptoms. And we try to apply a combination of approaches to give them answers. And eventually, for at least for a small subset, to be able to find specific therapies. And this network really grew out of an idea that started at the NIH Medical Center, where there was an undiagnosed diseases program founded by a man called Bill gall. And there was some impetus at the NIH to be able to spread the love, because they were absolutely overwhelmed with cases once the word got out. And so a group of us came together, and it really was a large, large group in building this network, which would then make that kind of medicine available much more broadly. And there's actually even an international version of the Undiagnosed Diseases Network now, and there's a network element. And then of course, there's a very individual element. And so of course there's one center for each patient or each family, but there's also the network element where you get to pick the brains of anyone in the network. And what I've loved about it, and the real privilege I feel in being involved, is that it combines the very best.
Dr. Euan Ashley: [00:25:01] I think at least we aspire to, to make it the very best of traditional medicine, where a certain number of cases it's something like 10 or 15%. So it's not a majority, but they're solved by literally talking to the patient, laying hands on the patient, thinking deeply about the patient, asking them about their history, really, the Sherlock Holmes sort of approach. Let's really dig in. Let's let's find the clues. Let's try to get the answer. And then the other part of that is let's go talk to anybody in the entire world who could help us solve this mystery. So some of it is very much that. And that's traditional medicine, like we all imagined, I think when we were younger, that maybe we'd spend most of our time doing things like that. Of course, most of the time we we have a much more protocolized approach to just get through the day and make sure we see as many patients as we can and help as many as we can. But in this case, we're we're spending many, many hours per patient, per family just trying to work out the answer. But it's the combination of that very traditional approach to medicine with the absolute cutting edge. And you mentioned you mentioned genetics. It's true that over these last ten years, maybe 20 years, we've gone from having sequenced one genome, the Human Genome Project, which was ten years, you know, billions of dollars, ten countries just to sequence one or actually about half of a genome.
Dr. Euan Ashley: [00:26:16] Now to the point where we can pretty routinely sequence genomes for patients with rare disease. And we do that on a daily basis. And so we can bring technologies like genome sequencing, not exclusively genome sequencing, but technologies that are on the edge of what is applicable in medicine. They're just coming out of the science labs. We can apply those tools to try and solve these conditions. And if we need to make stem cells to do that, that will do that. If we need to find another disease model to do that, we'll do that. We work with groups all over the country who can bring really any technology to bear to try and solve these different conditions. So it's a very rewarding group to be with and even to solve one case, it can be incredibly rewarding. But it's it's in a way the very opposite of what we're doing in the critical care unit, as we talked about earlier, where you're turning a dial and seeing an immediate change to someone's physiology, this is people who've waited years, perhaps to see you and eventually end up in the clinic, where we then apply over the course of months and sometimes even years, the absolute cutting edge of medical biotechnology to try to solve their case and bring them some answers.
Tyler Johnson: [00:27:22] Tell us one story. To paint a picture like someone who was the sort of prototypical had visited every doctor they could think of. Couldn't find an answer anywhere. And then not only were you able to help them, but you were able to really pinpoint here is the problem. And even better, if there was then something that could be done about it that actually helped them to feel better.
Dr. Euan Ashley: [00:27:43] Yeah. Of course. And one family who've been happy for us to share their story that came to our center at Stanford, but they came after several years of navigating a medical odyssey. I mean, this is named after, of course, the epic Greek poem, you know, a long and arduous journey. So in this case, there were a couple, Danny and Nicky, who had two boys. First of all, Carson was their first. And he started to progress from a developmental perspective normally. So if you put yourself in their shoes, here's a little baby bundle of joy and starts to progress. And as you if you've any of your listeners are parents, they'll remember every single milestone. The babies seem to change every week, or certainly every month. They can do new things. And he did that pretty normally through most of the first year of his life. But around year one, when most babies and kids by that point would be standing up, maybe cruising around the furniture, they could hold their own spoon, perhaps feed themselves. And he'd been doing all of those things, but he started to go backwards. They started, you know, taking him to a doctor, first of all, to a GP, and then from there to developmental specialists and from there to a neurologist. And the end I think they saw 3 or 4 neurologists. And despite the fact he was continuing to decline from a motor perspective, his intellect was pretty normally progressing.
Dr. Euan Ashley: [00:28:59] But from a motor perspective, he was going backwards. They had no idea. They said, well, we're going to call this cerebral palsy, but that's because we don't know what this is. And so in that period of time, his younger brother was born and he went through exactly the same trajectory. And so at the age of one, he started to decline from a motor perspective. And if you can imagine the lives that they had at this point, Danny was working full time, Nikki was full time on the kids. So she would be looking after each one of them, helping them get dressed, helping them get up, helping them get fed, you know, just exhausted by the day's end and the day's end. Danny would come back from work. He would take over. She would get some rest. He would then get them off to bed, and then he would spend the next 4 or 5 hours literally on the internet, you know, desperate to find an answer. And no doctor had been able to give them an answer. They'd even they'd had multiple tests. They'd had MRI scans which showed abnormalities in their basal ganglia, the motor part of their brain. So they were like, well, I think this might be the answer. This might be why they're suffering. But that was not really an answer. That was just really a glorified symptom, if you like.
Dr. Euan Ashley: [00:30:04] Is it from the perspective of a, you know, a scan didn't give any answers, didn't give any clues. And so every night he would be doing this. And, you know, just thinking somewhere on the internet he could type some keyword in that would take him to somewhere down some rabbit hole on the internet that could help his boys. And then eventually, after, I think, the fourth or fifth neurologist, they learned about the Undiagnosed Diseases Network and they had a referral into us. One of the things that we do is routinely sequence the genomes of all the individuals in the family. So each of the boys had individually had their exome sequenced the axioms, just the protein coding part. About 2% of the genome, and really nothing had been found. But what we did was sequence the entire genomes of all four. And the advantage of doing that is you can start to leverage the genome segments that are shared to really understand what it is. The two boys share that neither of the parents have. And what became clear very quickly when we did that was that both the boys had mutations in this particular gene, shortened to maker. Mitochondrial two enoyl CoA reductase. Bit of a mouthful. It's a gene that's critical to a pathway called it's a mitochondrial fatty acid synthesis pathway. So the little batteries inside cells of the mitochondria and they produce energy for the cell.
Dr. Euan Ashley: [00:31:21] But they also have their own metabolism. And this is a fatty acid lipid metabolism for those those mitochondria. The reason it hadn't been found before is that one of the variants was sort of in the in the upstream part of the gene. So it wasn't tested. But the real thing was that actually each boy had received one genetic variant from each parent. So the parents were both normal, Danny and Nicky both normal, but each boy had received one variant each, and it wasn't until we had the whole genomes of the whole family that we were able to see what this was, so they became the seventh and eighth boys in the world to be diagnosed with mi pan syndrome, which is the syndrome that only been in 6 or 7 others to that point. But here was the key. They they now had a name. They had a label. They had something, you know, and knowing what it is, we think as doctors, well, that's just half the story. If we can't treat it like, what's the point of knowing? But actually patients don't say that. Families find it incredibly empowering to go from what they call undiagnosed island to a place where they have an answer, because now they have they have a keyword like they can go to the internet and start looking this up.
Dr. Euan Ashley: [00:32:23] So that's what Danny did. So he started like typing in maker mitochondrial two enoyl CoA reductase, started a mini biochemistry tutorials. You know himself just learning about this. And he realized at one point that one of the things that is missing if you have that gene knocked out is lipoic acid. And so he started typing the downstream things in, wondering if there's any way he might be able to get hold of those and give them to his boys. And indeed, actually, rather than being a $3 million gene therapy. Lipoic acid is a supplement you can buy online, literally on Amazon. And so all of a sudden, his his nighttime searching on the internet led him to an Amazon site where you could buy something that would arrive in the mail the next day. And with the permission of his neurologist, of course, he started giving it to the boys and they started improving. They started to improve in their motor milestones. And it also since then he started a foundation. And now there's research looking for a genetic cure. So this is the sort of, you know, there's just one story, But just to give you an example of some of the things that we can do and just just how incredible it can be to be in the middle of the lives of these, these patients who go through such an incredible journey.
Henry Bair: [00:33:31] So you've been sharing a lot of really exciting and cutting edge technologies within medical genomics. I want to slightly pivot now to a very interrelated concept, which is precision medicine. As we know, you're also a leader in this realm of medicine. So can you tell us more about what that is?
Dr. Euan Ashley: [00:33:50] Yeah, it's kind of a buzzword in a way. And it arrived out of not nowhere because actually, President Obama first used the term precision medicine, at least to my knowledge. It's interesting, if you look at Google that Google searches for the term precision medicine went up straight after his state of the Union, where he mentioned it. Now, he wasn't the first person to coin it. Of course, at National Academy of Medicine, group had come together to think about new taxonomies of disease. So thinking about how we should classify disease for the new era. And they had defined this term precision medicine, and it was elevated then by President Obama and his speech. And I often joke that we all immediately went through our slide decks and changed, you know, personalized medicine to precision medicine for all our grants and all our papers. But what does it really mean? Well, you know, I think for some people, it really does mean the idea of personalized medicine that somehow something about your genome will help your doctor decide which of the high blood pressure medicines you should have. And we certainly have made some progress along that road. But what Obama meant, actually, when he used the phrase is something else, which is that we can now, because of molecular means, not solely molecular means, but often molecular means define diseases at a deeper level. And the example he used, which is sort of a classic example for this. And there are others I can talk about. But it was cystic fibrosis which is a genetic disease, recessive disease, obviously very challenging, where secretions from various organs, lungs and pancreas particularly are very thick and those mucus secretions lead to multi-system challenges.
Dr. Euan Ashley: [00:35:18] But pulmonary infections being a big one for those patients. And up to that point, very little targeted therapy had been available. And we thought of cystic fibrosis as one disease. But increasingly we've been doing genetic testing of those patients. And we realized that certain variants, genetic variants, were recurrent. And so there would be another way of classifying cystic fibrosis according to whether that particular genetic variant led to that channel as a chloride channel getting to the cellular membrane or not. And a company called vertex at the time had a drug that could increase the throughput of that channel. But of course, only if that channel made it to the surface. And so only a subset of patients benefited. But if you were in that subset, you got huge benefit. So about 10% of patients had this variant five, five one in the Cftr gene. And for those patients, this particular drug ivacaftor was tremendously powerful compared to what was there before. It was dramatic. The company then went on over the following years to say, well, about 90% of patients have this 508 deletion. And they had another follow up drug that was able to help stabilize the protein so it would traffic better to the surface so that drug would help those patients. And then various combinations now mean that more than 90% of those patients can be treated with something specific. So this was the vision. And actually Obama used this early example that a lot of that drug development has happened in the decades since. But that early example was the beginning of precision medicine. We take a disease, we understand it at the molecular level. We're able to subclassify it at the molecular level and then treat those individual groups with exactly the right therapy for them.
Dr. Euan Ashley: [00:36:53] And that's, you know, one example, you had Sue Desmond-hellmann on a little while back. I was listening to that episode, which was great. You were talking about her two breast cancer. You know, that's another great example where sub classifying a disease that was formerly thought about as a large group and then targeting it with a specific molecular medicine that is designed for for that. And in that case, it's not genetic testing, but rather a histology sample looking under the microscope. That would help us understand who should get that drug and who might not. And then in my own arena, maybe just one last example, you know, if people present to our clinics with very thick hearts, there's a number of diseases, many of them genetic, that can cause that. But, you know, one is fabry's disease, one is amyloidosis. You know, another is hypertrophic cardiomyopathy. And in the past, if especially if we look just with an ultrasound at the heart, we really couldn't tell which of those diseases, those molecular diseases were causing the hypertrophy. But now using genetic testing we can say, oh yeah, you have Fabry's disease. So you need enzyme replacement or you have amyloid disease. Now we have a whole raft of new medicines coming, silencers and stabilizers and others that can actually genuinely change the trajectory for those patients. Or you have hypertrophic cardiomyopathy. We now have a drug mavacamten that's a myosin inhibitor. So I think that's the vision and that's the excitement around precision medicine that we are increasingly able to subdivide diseases and then treat them specifically those subcategories.
Tyler Johnson: [00:38:18] Well, I think it's interesting too, because when you start to think about that, I you know, I think about this mostly from a cancer perspective, of course. Right. It's very interesting if you think about if you have 100 people who all present to an oncologist and all of them have cancer, and then you say, okay, now we're going to try to divvy these people up and decide who should get what kind of treatment. How should we divvy them up? The two ways that we have traditionally done that are to either say, okay, well, let's try to figure out where did the cancer start? Or we have taken a sample of the cancer and looked at it under the microscope. And then we've made these classifications. For example, if the cancer cells form what pathologists call glands under the microscope, then we classify that as an adenocarcinoma. Right. And if it looks a different way then it's a sarcoma or a squamous cell carcinoma or whatever. There's actually not necessarily it is not necessarily molecularly or biologically intuitive, viewed from a larger scope, that all cancers that start in the breast or the colon or what have you should necessarily be treated the same way, right? Right. And over the last 5 or 10 years, we have started to have the first ever tumor site of origin agnostic FDA approval that are based purely on molecular or genetic characteristics. Right. So now any solid tumor, regardless of where it starts, if there is microsatellite instability, which is a molecular finding in some cancers, then they can be treated with immunotherapy regardless of whether it started in the breast or the colon or the lung or where have you.
Tyler Johnson: [00:39:51] Right. Right. By the same token, tumors that overexpress Her2, which you're referring to a minute ago, regardless of where they started, they can be treated with a medication called an Her2. But the thing that's interesting to me about that is that it calls into question, to some degree, the entire way that we think about disease. By the same token, if you do a genetic test on someone when they are a child, because let's say that their parent had some kind of physiologic abnormality, and you find that they have the potential for an enzyme deficiency or whatever else, even though they may have no symptoms and may not even have findings that would be visible on a scan or an echocardiogram or what have you. In a sense, they have sort of a pre disease, right? Yeah. And the idea of being able to intervene at the stage of a pre disease is both sort of thrilling in the sense of increasing human flourishing, and yet at the same time also carries with it the shadow of something like Gattaca or, you know, something where everybody is sort of carries around their pre diagnoses and prospective employers or whatever, could know all of the things that might happen to you somewhere down the road. It becomes very quickly, philosophically and ethically complicated.
Dr. Euan Ashley: [00:40:59] Mhm. It's good that we do. I mean we have not every knows but we do, we do have a lot. It's not perfect but we do have the Genetic Information Nondiscrimination Act which does prevent employers and health insurance is discriminating against you on the basis of your genetic findings. Interestingly, that was in place before the preexisting conditions protections that came with the Obamacare act. And so for a while, you were protected against someone basically denying you health insurance because of a genetic cause, but not if you actually had the disease. So that was a strange moment in time. Of course, you know, we need to be in a world where none of that is true. But I think the great thing about you said, I mean, the emphasis is there is that most people will feel empowered by understanding as long as there's something that can be done. Now it's different matter. If there's a genetic finding for which there's really nothing that can be done, and people obviously should always be empowered to get the information they want and not be prevented from having information that they otherwise would want. But if there is a as for many diseases, something that can be done in prevention, then I think as a healthcare system and community, we should try and try to move more closely towards prevention, obviously, rather than response and reaction to diagnose disease if it's at all possible.
Henry Bair: [00:42:16] What do you think about widespread genetic testing of the general population? What are the limitations of that or the the benefits of that?
Dr. Euan Ashley: [00:42:25] Yeah, I mean, I think the first thing I would say is that if somebody wants their genetic information, then they should have it. You know, I think there has been quite a lot of medical paternalism around genetic information. So, you know, people shouldn't be trusted to have such data. I mean, it's your information. It's your genome. It's present in every cell in your body. If you want it, you should have it. And when we sequenced genomes of patients, we allow anyone who wants to have a hard drive with their genome on it to take it. Now, in Silicon Valley, there are indeed some people who take that home and analyze their genomes, but we're not a normal part of the world. I'm not sure I've spent really any time recommending that everybody gets their genome tested, but I think there are categories of patients where it's incredibly valuable. We've talked about some categories here on this podcast, and it's very frustrating that insurers have not recognized the clear benefits and cost effectiveness of doing that in those populations. And sometimes, in contrast with their willingness to pay for therapeutics, there's this somewhat reluctance to pay for diagnostics, which can be very frustrating. But your question was really more about the general population and other benefits in prevention. I think that overall there are benefits in prevention and some benefits that are increasingly held in what we call pharmacogenomics. The idea of using knowledge and understanding of your genome to help prescribers choose the right medication for you. And I think that if we could flick a switch and have that information available in each of our medical records, it's pretty clear the modeling clearly shows it that there would be fewer complications from prescribing.
Dr. Euan Ashley: [00:43:58] We would much more accurately reflecting the actual effect of the medications we're giving, because we'd have an understanding of how it was metabolized for an individual. So that would be one benefit. That would also be a small percentage, probably single digit percentage of people, especially if we we did their sequencing early in life in whom we could find something that will be important medically to their life and for which there could be a preventive therapy. So, you know, that won't be for the majority of people, but it will be for a significant minority of people. And then there are ways in which we can use things called polygenic risk scores to dial up or down a little bit. It's not terribly dramatic, but dial up or down a little bit. The risk of almost any disease for which we've done a genome wide association study, and it may not be worth it to do that for individual diseases. And certainly the jury's out a little bit on an individual disease basis. But when you think that that single test could give you that risk dial for any disease you wanted, then the cost of fitness element and the information flow element is a bit different. And I think that we've actually even struggled to properly model what that would look like. I think net benefit. But but you wouldn't find me. I think evangelizing this as the most important thing we should do in the healthcare system today.
Tyler Johnson: [00:45:10] One of the places where you have also lent some of your focus is to the effects of artificial intelligence on health care. And we've had a few guests on the program with whom we've delved pretty deeply on artificial intelligence in healthcare. I think that, you know, maybe what was it, nine months ago or something? The advent of ChatGPT for and some of the generative or close to generative AI models was was sort of the, you know, in large language learning models and whatever was sort of the that was all the buzz. Now the buzz has kind of faded into the background a little bit. We're not, you know, talking about it quite as actively as we once were, I think partly because some of the places where people have tried to implement that widely. Right. Like there was the the sort of infamous moment when Google tried to incorporate its AI algorithm into its search results, and it had some sometimes embarrassing results, even in pretty straightforward questions and some of that. And so I think some of the, some of the hype has died down a little bit.
Tyler Johnson: [00:46:09] But I think all of that makes this actually a really ripe time for asking the question, because I feel like on things like generative AI, on the one hand, you have some people who will make these incredibly bold predictions that this is going to, you know, just absolutely transform the face of healthcare. It's going to make many kinds of doctors all but superfluous. It's going to have these field changing dynamics and then other people who really Pooh Pooh that and say, well, no, it's really, you know, this is all hype and it's going to sort of fade into the background. It's really not going to be that much better than a Google search or what have you. So I'm curious, as someone who has some expertise and insight into that, how big of a deal do you think this is? Do you think we're like on a on the sort of the precipice of the dawn of a new medical era, or do you think that the the effects are going to be more muted than that?
Dr. Euan Ashley: [00:47:01] Well, I Generally considered, I think, an optimistic person. I'm kind of a glass half full sort of person. So let me say that up front. The dawn of the precipice of a new era, I think, is probably a little short of of where I would go. But I would say I'm very optimistic about the potential for this third revolution of AI. I mean, remember I was coined in the 50s and 60s, and maybe we had the AI winter in the 80s and 90s, and then we had the deep learning era where Google Photos could suddenly recognize our dog and our Frisbee and our cat. And then now I think of this as almost like the third revolution of AI. But I don't shy away from the term revolution. And I think the fact that it has dominated not the techie pages of our newspapers, but the front pages of our newspapers, is one of the testaments to the effect it has had on our lives and our culture, but specifically in medicine, obviously, that's a different matter. But what I think about first, when I think about the positive potential, is that not everyone has access to empathy. Not everyone has access to a doctor that can discuss their condition with them. And not everyone. Certainly not everyone, has access to an expert. In fact, most people have none of those, or at least the last two.
Dr. Euan Ashley: [00:48:16] I hope that maybe most people have some someone they can turn to, but there's a large number of people, really who have nobody. And while a computer and an artificial intelligence is not a substitute for a human, in my view, and I think most people's view, we do have a deficiency of of empathy and a deficiency of time for the doctors that we do have to be able to explain things in the way they would like. And then obviously a deficiency in expertise. People who really know, especially with rare condition, you may have to travel halfway around the country to find somebody who's really an expert. So what if we could use these artificial intelligence models to extend those doctors, those physicians to extend that expertise, to take the best of medicine that is available wherever it is in the world, and democratize it, make it available to many more people. I think that's the promise of the current AI revolution, and I don't think it's unrealistic. I mean, there are some data to support it. Obviously, in terms of expertise, the models have progressed their way through the medical board exams, the law board exams, the education board exams from an average score to an expert score. They've now taken on the rare disease cases from the New England Journal.
Dr. Euan Ashley: [00:49:28] And they certainly equal our ability to make those diagnoses. And I know you've discussed this before, but they also are judged by not just patients, but also other doctors as being more empathetic in certain situations. Um, now, I think one of the reasons that judges are more empathetic is that they have, in essence, infinite time to write the response. I mean, they work infinitely quickly, if not infinitely quickly, but they work so quickly that they can generate much more data, many more words, and appear much more empathetic because they would then have more, appear to have more time. And certainly, you know, I have patients and I have folks who work in my science lab that have GPT or its equivalent on all day. I mean, it helps them with coding. If we think about the science lab, but it also they talk to it. I do it myself like I have it. At least one GPT or equivalent version open all the time. If I'm writing, it's a helpful kind of thesaurus aid. It's a little better than a thesaurus. It answers questions better often than than Google these days. But of course, that brings us to the the peril of AI. And nobody should discuss the potential optimistic scenarios I just outlined without worrying about the elements that we all spend time worrying about, not least the obvious ones that it often because it's a word generator in probability terms, it often just makes things up and gets things wrong.
Dr. Euan Ashley: [00:50:51] And that is unacceptable. Although of course there are many doctors who sometimes make things up and get things wrong. But I don't think we approve of that in general. So, you know, I think that is a challenge. I'm certainly not ready at this point in time to leave one alone with one of my patients and then walk away. But would I be willing to leave one alone with one of my patients and then step in at the end, the way we do with some of our fellows and residents to say, okay, let's talk about what you've been discussing with the system. If that system was based on, let's say, my practice and my words and my writing in a way that would allow me to see more patients, maybe. I think that's that's an interesting question. But we still have some unanswered issues as well as accuracy is one, confabulation is another. And of course, the amplification of the bias that exists in the training data is probably the biggest one. So I'm very much a glass half full person. I do think there's a lot of opportunity here, but also clearly we have to spend some time thinking about the downside as well.
Henry Bair: [00:51:49] So I'm currently working well. I'm currently an ophthalmology resident, but last academic year I was an internal medicine intern working at Thomas Jefferson University Hospital in downtown Philadelphia. And the patient population there is very much it's urban. It kind of feels like a community hospital. We take care of lots of unsheltered patients coming off the streets, and it was such a drastic change coming from Stanford because not just not necessarily just the patient population, but also the way that the culture, the medical culture there thought about innovation and AI and precision medicine. Because, I mean, even now, like, I still get I still, you know, I'm still I still have my Stanford like, email active. I still get notifications, announcements from Stanford Medicine about all these events and conferences and lecture series about the cutting edge. Right. There's there's so many things about design thinking and healthcare innovation and all these conference series featuring the leading thinkers from private industry and from academia, and that just doesn't exist at the place where I did my medical residency. And frankly, all those things don't exist in most places around the country. The culture at Stanford is not representative of how most hospitals around the country, around the world runs. So, you know, you pointed out a lot of limitations and pitfalls of how AI currently works and how genetic medicine is deployed. Absolutely. But let's say that it all works well. How then, do you think about the equitable access and distribution of all of these powerful tools?
Dr. Euan Ashley: [00:53:34] Yeah, it's such an important question. And like so many things that are new and more expensive when they're new and where there's a learning curve for everybody in terms of using them and that they start in a small number of select places, but we hope they eventually reach everybody. Well, I think in two ways. We talk about genomic medicine. I think the good news is that sequencing is so much cheaper. I mentioned the Human Genome Project earlier, $3 billion. I mean, if you sequence enough genomes, you can now see the genome for 1 or $200, which is a remarkable change, which means that if somebody wants that information, it's close to the cost of an x ray. So I do think that that has happened because of the work of many. And although it has been available to a select few in those first years, it is now available much more broadly. I think the difference a little bit with AI is that I think it could be deployed early to spread that expertise. While compute certainly is not cheap, I mean, there's a reason Nvidia has become $1 trillion company, and it's because compute is not cheap. But assuming that compute can be made available, and remember that the compute is really for training the models, actually deploying the models has a relatively much less of an overhead in terms of compute. So they can be deployed locally and even on phones and maybe one day on watches. And so I think that can actually play a role in, although it seems like something that would only be available to a few, I think it can play a role in being available to many.
Dr. Euan Ashley: [00:55:04] And certainly all of these large language models right now have a free option where you can just literally go to the website and log in and start asking questions. Now, that doesn't mean I'm recommending I'm not recommending anyone go and start asking medical questions that way. I think you have to be careful of of the sources, but but I'm not naive either. I people are definitely going to these models and typing in questions. And compared to Google when you might get, you know, the Stanford Medicine website or Mayo Clinic or Cleveland Clinic or something. You certainly know that that's going to be accurate data, which is not something you can say for sure when you type into GPT or cloud or one of the other equivalents. What you'll get is a very friendly answer, and you'll get a conversation that you can then interact with. And it is mostly correct for many medical things. But the problem is you can't tell when it's not correct. And so nobody, I think, should be making medical decisions based on off the shelf large language models. But as tools that add to the many tools we have for information in the world, they're currently being used and they're currently available to anyone with a phone or an access to the internet. And so I'm actually hopeful. I'm a technophile, obviously, but but I'm also hopeful that technology deployed in the right way can actually improve equity instead of certainly there are examples where it can make it worse and exacerbate equitable challenges. But I'm an optimist that it could actually improve our quest for equity and opportunity in healthcare.
Tyler Johnson: [00:56:37] I think I'm a little you know, if you've heard some of our other episodes, you've probably gathered that I'm a little bit more of a Luddite and a little bit more cynical about the good that AI will do. Not that it can't. It's not that I don't think it can do good, but I guess I'm not yet convinced that it necessarily will do good. I think it may do good is the way that I would sort of come down. But all that said, for you, as a self-proclaimed technophile and optimist, I'm genuinely curious. Just as a not that this particular question is the most important or pressing one, but I'm just as a sort of representative of a larger philosophical question. I'm genuinely curious if you or someone very close to you, someone that you love, let's say, was going through some really terrible thing, heaven forbid. Right? That the person you loved had lost a loved one, or they were just terribly depressed or what have you. And in effect, let's just pretend, for the purposes of this thought experiment, that you could wave a magic wand. And no, they didn't need an antidepressant. They didn't need an antipsychotic. They didn't need an admission to the hospital, psychiatric ward, or what have you.
Tyler Johnson: [00:57:47] They just needed someone to talk to. They just talk. Therapy was let's again say you could wave a magic wand and no, that was the best thing for them. And then let's say, you know, you could call some whatever the psychiatry referral number in your insurance network or something, and that in some future world the person said, well, we can either get you a referral to a human psychiatrist or psychologist who will see you in a couple of weeks, or we can text you a link to a bot that will, you know, quote unquote, have a conversation with you or chat, you know, do a chat conversation with you, or even maybe in some future world, a seemingly live some sort of droid version of a conversation. Let's say that you have that data that you referred to earlier that's a little bit famous. Now about that. Computers were scored as being quote unquote more empathetic than humans. Would you be okay with your loved one? And let's say you had to choose between the two. You can't have both. Would you be okay with your loved one being referred to the bot, as opposed to being referred to the real person.
Dr. Euan Ashley: [00:58:50] I think yes is the simple answer. No. You know, I think because, you know, two weeks is a long time, and especially if you're feeling really bad and really lonely, you know, people, it has been shown that they anthropomorphize these interactions. They even although they know it's a computer and we can feel it ourselves. I mean, I do it, I talk to this thing, I say, thank you. Thank you. Thanks for answering. I'm like, this is a you know, there's a bunch of ones and zeros. And I'm saying thank you. But it feels like the right thing to do because it appears like a conversation. And now I don't want to leap from that to, okay, this is now a board certified psychiatric crisis. Obviously. Obviously not. But when the when the question which is what you're asking is for today, the answer is you can have nothing or you can have that and that that is increasingly safe. And increasingly, you know, the guardrails at least have been significantly improved from the first version, I think I would not have said yes if it was GPT 3 or 3.5. The guardrails have changed significantly, and they're still not perfect. But I know people who have conversations with these language models and find them fulfilling conversations, even though their brain absolutely knows that this is a computer. And I think I can I can believe that. And also, it's a big leap to go from people marking. Yeah, that answer I find more empathetic to. I would rather I'm happy to talk now to a computer about my inner self, but some people, it would appear, are not only just fine with that, but likely have some benefit. And I don't want to. We really don't have the data yet to know that for sure. Sure. But your question was about me personally, and I think my answer is yes.
Tyler Johnson: [01:00:33] Fair enough.
Dr. Euan Ashley: [01:00:34] What's your answer?
Tyler Johnson: [01:00:36] I don't know. I so I guess that I have a functional and a philosophical answer. The functional answer is sure if option A realistically right now tonight or something is sit in your room and stew in your juices and have no outlet for your feelings. And option B is a bot who you know, appears in some synthetic way to be empathetic. Sure, the bot is probably better than stewing in your juices alone in your room at night or whatever, right? But I mean, the bigger, of course, philosophical question, which is the, you know, and of course, this is kind of the grand question of AI, right? Is, you know, similar to, I guess, a different facet of the whole question of when does an AI become sentient or really start to think, or when does it become a truly general, generalizable artificial intelligence or whatever? I think a related question to that is forget about board certification and the kind of I don't mean to diminish the work of psychiatrists, but the, for lack of a better word, the sort of technocratic dimension of whatever it is that's involved in talk therapy or what have you. But I think the bigger question is, is there Is something essential that talking to a human has, that talking to a bot. I don't think the biggest question is whether the bot can have it right now, but is there some essential thing about talking to a human that a bot can never have? That no matter how advanced, no matter how many words it has processed or how many, you know, iterations of large language learning models it's been through or whatever, is there some essential element of a connection between two humans that just cannot be replicated in a bot? And then I guess another question from a medical perspective would be, could you capture such a thing in a clinical trial or whatever? And I guess if I'm being totally honest, the only thing that I can say is that I hope that the answer to that question is yes. I'm not sure that I know that the answer is yes, but I hope that it is.
Dr. Euan Ashley: [01:02:35] I'm completely in agreement with you. I mean, I was going to say it's the same thing. I hope so it would be my answer to that. I don't know either. I can't say for sure. What I know is that these models are very good at doing the things that we teach our students to do, you know, in terms of reflecting the emotions of the person who's talking to you. They do that very well, and it turns out you can get quite far with that, I think is sort of what I've learned in interacting with them. But I really hope that there's a space left for us humans, whether it's in better understanding the three dimensional element of interacting with another human and really having a lived human life, rather than being a series of of words that are probabilistically laid in front of one another.
Tyler Johnson: [01:03:20] And to that point, I've read a number of sort of thinking analytical pieces about the AI revolution that I guess I just don't have the conceptual knowledge base enough to know who I agree with, because there are some people who at least, you know, have whatever the credentials to be able to say that they understand this at a really deep level, who say that the entire idea that an AI could for example, demonstrate empathy is misplaced because an I. As you said, all it's doing is using incredibly sophisticated probabilistic models to predict what would humans, en masse, consider to be the most appropriate next string of words, right? Like or what is the most likely thing that they would say that someone in this position should say next, but that that's entirely different from a person actually seeing your pain and appreciating that you are suffering and being there and sort of being present with you and witnessing that. I guess a separate question is whether an even if an AI can't do any of those things, can it act in a way that is so superficially similar to a person who is feeling those things that it becomes indistinguishable to a human, which I guess is a very sophisticated version of the Turing test.
Dr. Euan Ashley: [01:04:34] But yeah. No, I mean, the question, does it matter in the end if it can appear empathetic in a way that makes you feel better? Does it matter if it never experienced your pain? I. You know that. I think these are important questions, actually, that I'm. You know, I'm glad that we're discussing them.
Tyler Johnson: [01:04:48] I, I guess I, I want to pound my fist on the desk and say, yes, it matters. I'm not even sure I know why it matters, but I really want it to matter. That's what I can say. Okay, we've been way over time. So the last question we want to ask you, which we ask all of our guests, is if you could go back to well, and you work with people who are in training all the time, but if you could distill down a thing or two that you feel like this is the thing that you, as a health care professional in training, need to make sure you know and never forget. What would you say?
Dr. Euan Ashley: [01:05:20] I think never to forget that the patient in front of you is also the human who is out throwing a ball with his kids yesterday, or fishing with his grandkids last weekend. I often will recommend to patients and family members and residents that if we can get pictures from the patient of their life outside the hospital that we put them up in the room, or we put it on the iPad and make sure it's there, because I think that it's and I find myself doing it too. It's so easy to forget. And actually, one of the things Covid did was remind us that people have these lives. We saw them in their homes. A lot of the time it's hard to choose one thing, but if I had to choose one thing, it would be that, you know, as an aid to remembering that the person in front of you in a hospital bed is also that person out there having a life that's been really profound for me. When I go into a patient's room and see them, how they exist, when they look well, and that's how they look to their family.
Tyler Johnson: [01:06:17] And I think in a funny way, that is your own way of putting your nickel down and saying that no, humans cannot ultimately be reduced to their own biological versions of ones and zeros, right? That there is some ineffable essence of being human that matters beyond just how their organs are functioning 100%. Well, thank you so much, Doctor Ashley. We really, really appreciate all of your good work and your good thoughts and that you would take the time to be here with us.
Dr. Euan Ashley: [01:06:44] Yeah. Thanks so much for having me. I really loved the conversation and thanks for doing the podcast. Keep up the great work.
Henry Bair: [01:06:49] Thank you very much, Euan. Thank you for joining our conversation on this week's episode of The Doctor's Art. You can find program notes and transcripts of all episodes at the Doctor's Art.com. If you enjoyed the episode, please subscribe, rate and review our show available for free on Spotify, Apple Podcasts, or wherever you get your podcasts.
Tyler Johnson: [01:07:12] We also encourage you to share the podcast with any friends or colleagues who you think might enjoy the program. And if you know of a doctor, patient, or anyone working in healthcare who would love to explore meaning in medicine with us on the show, feel free to leave a suggestion in the comments.
Henry Bair: [01:07:26] I'm Henry Bair
Tyler Johnson: [01:07:27] and I'm Tyler Johnson. We hope you can join us next time. Until then, be well.