The future of coronary heart disease

0:00Mike McConnell:This is Stanford Engineering's The Future of Everything, and I'm your host, Russ Altman. I thought it would be good to revisit the original intent of this show. In 2017, when we started, we wanted to create a forum to dive into and discuss the motivations and the research that my colleagues do across the campus in science, technology, engineering, medicine, and other topics. Stanford University and all universities, for the most part, have a long history of doing important work that impacts the world. And it's a joy to share with you how this work is motivated by humans who are working hard to create a better future for everybody.

0:37Mike McConnell:In that spirit, I hope you will walk away from every episode with a deeper understanding of the work that's in progress here, and that you'll share it with your friends, family, neighbors, co-workers as well. The cancer world has really pushed the idea that if we can detect things early, treat it aggressively, we can cure cancer. And I want people to think about heart disease the same way.

1:05Mike McConnell:This is Stanford Engineering's The Future of Everything, and I'm your host, Russ Altman. If this podcast brings you some value, please rate and review it. That's a great way to spread news about the podcast and grow our listenership and also improve the show through your comments in the reviews. Your input is extremely valuable, so thanks for doing that. Today, Mike McConnell will tell us that you want to learn about your heart disease as soon as possible so you can make changes, get medications, and live a long, healthy life. It's the future of coronary heart disease. Today, we're going to continue our feature called The Future in a Minute.

1:37Mike McConnell:I will ask Mike a few rapid-fire questions. He'll give me his best rapid-fire answers, and it will be fun. Before we get started, don't forget to rate and review. We love those.

1:54Mike McConnell:So heart disease, it causes heart attacks. You don't want to get a heart attack. Many people, it's the number one killer in the world, get heart attacks. So we know that there are great treatments. We know we can take statin medications to reduce cholesterol, and there are some amazing other new medications. We know that you can exercise, you can make changes in your lifestyle to prolong your life. But what about finding out that you even have heart disease? You don't want to wait until you're having chest pain whenever you exert yourself. That means it's far advanced. So there is a new sense among cardiologists that we have to do early detection of heart disease.

2:30Mike McConnell:And in fact, new technologies, including AI, have enabled amazing new ways to detect the very start of heart disease so you can intervene quickly. Well, Mike McConnell is a professor of cardiology at Stanford University and an expert on early detection and prevention of heart disease. He's going to tell us that these new methods of detection really are mandatory so people can find out if they have a disease and intervene. His hope for the future? Way less heart attacks and longer lives. Mike, you wrote a book called Fight Heart Disease Like Cancer. Can you explain that analogy and why it's so important?

3:10Mike McConnell:Thanks, Russ. Yeah, it would be great if there was a really short answer to that, but... We got 25 minutes. Okay, we've got a good opportunity to talk about it. Really, it came from a lot of experience working with patients and also in circumstances around my family, realizing that as much as we recognize heart disease as leading cause of, you know, death and disability around the world, we still view it as thinking, you know, we can just sort of do lots of prevention, don't really need to think of it around the way we do cancer, where we really need to do early detection and thinking about treating it early.

3:56So we never have to get to a point where people are having heart attacks and strokes. So there's really many aspects, right? One, it is like cancer, a biologic disease with these small growths that start in our blood vessels. You know, the original term, you know, for these growths was atheroma, which is similar to, you know, the OMA that we think of carcinoma. Yeah. And so it's really getting people to reframe how they think about it, you know, especially around the early detection. And then when you find it, you know, we now have lots of therapies that we can really stop it in its tracks. And, you know, whether you want to get into the words of eradicate or ultimately cure the disease, you know, the cancer world has really pushed the idea that if we can detect things early, treat it aggressively, we can cure cancer.

4:52And I want people to think about heart disease the same way.

4:55Mike McConnell:Great, great. So yeah, so that's really helpful. And one of the themes you just kind of alluded to is that there should be a sense of urgency and that sense of urgency needs to be early, not when the disease is advanced. So just for people who don't think about heart attacks or coronary artery disease very much. You already mentioned that it's related to the buildup of what we call plaques or cholesterol glops that are blocking the blood flow in your heart. Could you quickly review the risk factors that people should worry about that you as a cardiologist are always asking about? And also, how is the diagnosis made?

5:32Mike McConnell:Because I just want to set that up as a background and maybe the traditional ways for risk factors and the traditional ways for diagnosis? Because I know that you've been involved in a lot of really new ways to do that, which we can get to in a minute. Yeah, probably the most important thing to start is heart disease actually starts in our blood vessels, not in the heart muscle itself. And we have these, what are actually quite small coronary arteries that, you know, feed the heart muscle. And they're only about a tenth of an inch in size. And that's why a little bit of this plaque buildup, these atheroma growths inside the heart arteries, you know, if they cause a severe blockage or the most common situation with heart attacks is a blood clot, you know, forms on one of these plaques, blocks blood flow, and that's, you can have a heart attack or sometimes your heart stops.

6:25And, you know, a lot of people, their first symptom can be that they have what we call sudden death or cardiac arrest. And so, right, the risk factors people probably are aware of, I call some of them sort of the big three growth factors. So high blood pressure, high cholesterol, and high blood sugar or diabetes. But the Heart Association has also promoted the concept of life's essential eight. So the other things beyond those big three are physical activity, healthy diet, not smoking, getting good sleep, and maintaining a good body weight. So those are the sort of essential eight to prevention.

7:15And then you asked about diagnosis.

7:17Mike McConnell:Yeah. And maybe the traditional ways, because I know you're doing a lot of really new stuff if that's exciting. Yeah. Well, there has been a technique that's been around actually for quite a while that obviously AI and other tools are empowering even more. But the most simple and what's in the guidelines is when these plaques form in your heart arteries, they can calcify as they age. And x-rays are very sensitive to calcium deposits, just like a mammogram can pick up calcium deposits in a breast tumor, calcium deposits form in these plaques that form in our blood vessels. So a simple low radiation dose CT scan that takes 30 seconds can pick up these early deposits.

8:09And so that this corner artery calcium scan is the one that's been around the longest in the guidelines. And we can start getting into how AI is sort of empowering that even more. But basically, you're looking... The older way to do it was you were looking if you had a blockage that was really narrowing the blood vessels substantially. But part of this discussion is we want to reframe. That's really late. That's like looking for an end-stage tumor that's already causing lots of problems that's harder to treat. We want to catch that early, just like a mammogram catches breast cancer early when it's most treatable and curable.

8:57Yeah.

8:58Mike McConnell:So a lot of people are aware of these things that they call like the treadmill test, where, you know, you put somebody on a treadmill, you work them out, and you're checking their EKG and, of course, asking them if they're having any chest pain. I take it that that's late in the game. and does that still have a role? So for evaluating patients who have chest pain, stress testing, as you're describing, is still used. I do highlight in the book that the idea of using it as a screening test to sort of make sure everything's okay, I think is very misguided because one is it's only going to detect late stage disease.

9:42And two, people walk away if their stress test is normal. They could actually have a lot of plaque buildup in their heart arteries that just hasn't caused enough narrowing for the abnormal stress test. And so they think everything's fine. And you hear about people that had a stress test one day and heart attack the next. then that's because you don't actually need a severe narrowing to have a heart attack. And so that's why going directly more to the source of looking at the blood vessels, picking up those earlier changes is critical. Great.

10:19Mike McConnell:So now let's get into all the new stuff. And you said the word AI or the letters AI at least two or three times already. So why don't we go right to AI? How is that revolutionizing or is it revolutionizing these kinds of diagnostic tests and the ways that you can then approach the disease? Yeah. So one of the most recent examples, which actually happened sort of after the book, is some other colleagues at Stanford have used AI. When people had a chest CT scan for another reason, they found that an AI developed at Stanford could analyze that chest CT, see if they had those calcium deposits. And then when patients and their doctors were notified that these prior CT scans had the evidence of early heart disease, you know, over 50 % of them got on good preventive therapy within six months.

11:20And, you know, over 90 % stayed on it.

11:21Mike McConnell:I think you've called this, sorry to interrupt, I think you've called this opportunistic because they got the chest CT maybe for a pneumonia or whatever. But then they're getting this bonus finding, which actually might be more important than the reason they checked the CT in the first place. Yeah. So this concept, yeah, this opportunistic imaging that many pictures that we get of different parts of our body have a lot of health information beyond whatever the doctor ordered it for. Right. So the the other one where I've been working with a startup company is doing AI retinal imaging, something actually we first published back when I was at Google.

12:04Now it's almost eight years ago. It shows you can take a long time to get from a publication into a product. But similarly, when someone gets a routine photo of the back of their eye, it has a lot of blood vessels there. And so, you know, we can run AI to detect that you're at increased risk and really need, you know, a more thorough evaluation for early disease and likely need for preventive therapy.

12:32Mike McConnell:Okay. So that's, I'm going to stop there. And I want to dive a little bit into that because I think people will be thinking, wait a minute, my eyeball is telling you about my heart. So can you go through the kinds of changes? I know you wrote a paper about, I think, detecting hypertension by looking at the retina. I don't know if that was before or after heart disease, but you obviously have a great interest in using the eye as a proxy for the heart. So can you tell us what does the AI actually see that allows it to make these inferences? Well, you probably know that some of the things that AI sees we understand and some of the things AI sees we don't understand.

13:10Actually, the original publication when I was working at Google, it really blew away the ophthalmology world because the AI, they basically trained it against all these different risk factors, which include age and sex. And the AI was over 90 % accurate at telling men from women just from, you know, being trained on images of thousands of men versus thousands of women. And ophthalmologists who've been looking at retinal images their whole career couldn't, you know, didn't really, you know, this is sort of, there's a black box element and then there's some more explainable elements. Certainly, you know, when we did these, what they call these attention maps, you highlight the pixels in the image where the AI is basing a lot of its prediction.

14:11And certainly when it's trained on blood pressure, the pixels all light up along the blood vessels. And we've actually known for 60 years that cardiovascular disease causes narrowing of the arteries and widening of the veins, which you can see in the microvessels of your retina. So there are explainable aspects, you know, sort of the damage that comes from hypertension, diabetes, cholesterol. But there's also some things that we still haven't fully been able to explain of how the AI can pick up often a lot of early features that humans can't easily see.

14:55Mike McConnell:that's just that's really just amazing one quick question as a follow-up is um are the people who are having high scores for coronary artery disease risk from their retinal scan are they having visual problems or is this all below the level of detection of visual issues yeah so this is all below the level of detection of visual issues so and that's why on the other hand they're you know, it's estimated there are about 100 million people who go for eye exams because they do, you know, need glasses and other things, actually more than get sort of a regular primary care annual exam. And so that becomes this huge opportunity when they're going for their regular eye exam to, you know, sort of get flagged that you're likely at increased risk for heart disease.

15:49And it's good to know early and talk to your doctor about getting further checked and on preventive therapy.

15:57Mike McConnell:So now we've talked about two kind of amazing new ways. You talked about the calcium in the coronary from a CT. You also talked about the retina. You implied that the coronary calcium is going to show up early and so could be an early. Is the retinal changes also showing up earlier than other traditional markers of disease? Or is the jury still out on that? That's a good question. There are publications and other efforts that actually compare the retinal changes and they train those against calcium scores. Oh, good. In some ways, those two worlds may ultimately come together. Um, you know, I think a lot of what I've been interested in is how do we better identify when somebody needs a calcium scan?

16:51And so, you know, the retina potentially becomes a way to identify that we have a number of risk scores, but if you just look, I mean, when I worked on the book, one of the things that sort of blew me away was, even though I've been doing this for a long time, you know, when a man hits 50, about one in three already will have, you know, calcium deposits in their heart arteries. And a woman when hit 60, same thing, one in three. So, you know, this notion that you can just sort of look at somebody and say they look healthy and, you know, you don't have to worry about heart disease. You know, I sort of jokingly say, you know, I've been doing this for over 20 years, but I don't have x-ray vision.

17:41And, you know, so people that I think are at increased risk, some percentage will still have a zero score and some will have a score off the charts. And so we can much more personalize. And again, some other colleagues had published, which was a nice link to sort of this cancer approach is they took, you know, your calcium score could then stage your heart disease. So the same concept of cancer diagnosis and treatment. And so they, you know, stage zero through four. And then the intensity of the therapy matches the stage of your disease. So part of what I think is a good paradigm shift that's starting to happen in how we approach heart disease.

18:33Mike McConnell:This is The Future of Everything with Russ Altman. We'll have more with Mike McConnell next.

18:47Mike McConnell:Welcome back to The Future of Everything. I'm Russ Altman, and I'm speaking with Mike McConnell from Stanford University. In the first segment, Mike told us about the amazing new ways that we can detect coronary artery disease very early to maximize the chance of intervening and living a long life without heart attacks. In this segment, Mike's going to tell us about his excitement about direct-to-consumer technologies, wearables, AI on your medical record, that can also help with this task of early detection. To direct-to-consumer is tricky, though, because it means that you're taking some of the influence away from the clinician and putting it in the hands of the patient.

19:26Mike McConnell:That might not be a bad thing. Don't forget, at the end of the episode, we'll have our future in a minute segment where I'll ask Mike some questions and he'll give me some answers. Mike, you've been very involved in thinking about how we can empower a patient to kind of take control of their heart disease because their doctors may or may not have a sense of urgency that you think should be had. So tell me about why you're thinking about direct-to-consumer and how it's manifesting itself in the diagnosis of heart disease. Yeah, you know, a lot of what I've been trying to do over the last 10 years is How do we do this prevention at scale?

20:03And we recognize our healthcare system is super busy taking care of all of the acute active problems we had. And it's, you know, it's hard to spend a lot of time on the prevention side. Obviously, as a preventive cardiologist, that's the primary thing I do. But, you know, most people don't have easy access to preventive cardiologists all the time. And, you know, I've been excited over the last few years how prevention has really come into growing importance, right? Our current spending on healthcare is skyrocketing. I think there's a broad interest in how do we do more prevention. And I think in particular, patients or consumers, as part of the reason I worked in consumer companies like Google, is, you know, how do you reach many more people and make them aware of this?

20:58And, you know, obviously it's in, we're often the most interested person in our own health. And I think we see this in a lot of the emergence of what are now called sort of longevity companies, which it's, you know, it's exciting. You know, it's a new buzzword. It's ultimately how do we do better prevention? How would we have a better, you know, health span, right? How do we stay healthy, active, vigorous for longer? And everybody wants that. And now with, you know, we can have health records on our phones. We can, you know, there are multiple companies now offering to connect their AI agent to your phone health records to say, we can help you analyze it.

21:44We can find evidence of early detection. We can connect your mobile wearable data to your health record data, potentially to your genomic data. Um, so, you know, the power of this mobile ecosystem that pretty much everyone has a phone and, you know, the power of, of AI on top of it, I think we have a lot more opportunities to help people, you know, be more, uh, aware of and get more help. You know, sort of the thing I put in the book is, you know, everybody deserves to have a little preventive cardiologist sitting in their phone, monitoring their wearable data and their health records and trying to help keep you healthy.

22:33Mike McConnell:So it is remarkable that you say all this because really you're a cardiologist, you've been trained in the traditional way, right? And so you could have come on here and you could have said, I'm very worried about all of these pseudo cardiologists that, well, forget about taking away your business. I'm sure you're very busy anyway, but like maybe they're not doing a good job. So implicit in your comments is that you must think that these really have promise not to mislead patients, but to actually lead the patients to kind of better outcomes. So where are we? And like, what are the technologies that you think are the most like mature in the direct to consumer space where you would tell a loved one or a friend, yeah, doing that is not crazy.

23:15Mike McConnell:And there's actually some science behind that. Yeah, I think, right, you're you heard sort the optimist side of me. There's clearly the challenge of many businesses, you know, they focus on how to make money and to, you know, often promise things that sort of go beyond the science or the guidelines. And so that is the worrisome part, right? And many of the programs can be quite expensive. So they can actually make disparities in healthcare in many ways worse. I mean, right. So there's a company that I help that's very focused on South Asian heart health, because as a group, they have a number of risk factors and increased rates of heart attacks.

24:08And we intentionally work with the Heart Association and their, what they call their cardio, kidney, metabolic health paradigm. So what has me optimistic is, you know, many of the medical societies and healthcare system, you know, is trying to still contribute to this space so that we try to make it as guideline based as possible. You know, that life's essential eight that we talked about at the beginning, you know, blood pressure, blood sugar, blood cholesterol, You don't need a battery of 100 tests. You really, you know, if you can focus on the fundamentals, that covers 90 % of it. So there are risks in sort of over-treatment, over-testing, over-promising.

25:02Right. But I think the part that makes me optimistic is it's putting sort of prevention and health span, you know, much more into the public realm. and hopefully help many more people with early recognition of health conditions and engage them in all the things that they can do for prevention.