Longevity & Aging Series (S2, E4): Dr. Meesha Dogan
Dr. Meesha Dogan, CEO of Cardio Diagnostics, joins host Dr. Evgeniy Galimov to discuss her pioneering work in artificial intelligence, epigenetics, and genetics in developing next-generation DNA-based tests for preventing and managing cardiovascular disease globally.
Transcript
Listen to an audio version of this episode
Welcome to Longevity & Aging Series. Today, I’m going to introduce to you Dr. Meesha Dogan. Meesha has a track record of translating cutting-edge research into practical healthcare solutions.
Dr. Dogan is the CEO and co-founder of Cardio Diagnostics. She has over 10 years of experience in bridging medicine engineering artificial intelligence towards building solutions to improve cardiovascular disease prevention. With Meesha today, we are going to discuss cardiovascular prevention, how epigenetics can be used to improve it, and how the new tools developed by Cardio Diagnostics are related to longevity.
So, Meesha, over to you. Please tell us more about your company, about the problem your company solves, and why epigenetics is important in predicting disease risks.
Meesha Dogan
Well, first and foremost, thank you so much for having me. And I’m thrilled to be here. It’s lovely to meet you and to be part of your platform. So as you shared, I am one of the founders and the CEO of Cardio Diagnostics. I’m actually one of the inventors of the technology behind the company as well.
So at Cardio Diagnostics, we’re a precision cardiovascular medicine company. What that means is we bring together four key areas, which I would say are data, technology, insights, and values. So we use data, high quality, robust data, pretty large data in the form of molecular data, clinical data. You touched about epigenetics. That’s one element of what we do. We’ve built a one-of-a-kind technology for heart disease, specifically our AI-driven integrated genetic epigenetic engines, so our technology is actually at the intersection of those three things: epigenetics, genetics, and artificial intelligence. And we see ourselves as providing insights to various healthcare stakeholders, both clinically and non-clinically.
So if you’re thinking clinically, the more traditional provider-patient side of things. If you’re thinking non-clinical, groups like payers or even employers who are thinking about innovative benefits for their employees. And then, of course, ultimately, our goal is to drive value, whether it’s business value, whether it’s clinical and health outcomes in different ways focused on cardiovascular disease. So really, the reason we founded the company is if we think about heart disease in general, it has been the leading killer for over a hundred years at this point. We’ve made some incremental improvements in going from relying on things like just risk factors, like whether it’s cholesterol or diabetes, which we know are necessary, but they’re really not sufficient.
And then we started getting into more advanced technologies. And where we found ourselves is the use of epigenetics and artificial intelligence. And the reason we were able to get in the forefront of that is because I’ve spent, at this point, over 16 years in epigenetics and over 14 years in artificial intelligence. So really, one of the pioneers in the early on use of it, which gave us that added advantage to look at epigenetics from the context of the disease and how can we go to the molecular level of heart disease to understand the underpinnings of the development of the disease, but also use that as a way to inform personalized treatment, understand the effectiveness of treatment and ongoing patient monitoring.
Evgeniy Galimov
And can you tell us more about the products? So as I understand you have two main products, PrecisionCHD and Epi+Gen CHD, right?
Meesha Dogan
Yeah, absolutely. So we have two products as you shared. Both of them are clinical blood tests. So we are a clinical company, both of our blood tests are prescribed by providers. Happy to talk about how we’re also thinking of healthcare delivery because it’s one thing to develop innovative products, it’s another thing to make sure you’re getting it to as many people as possible, so scalability and accessibility side of things. But it uses the same underlying technology I talked about, which is genetics, epigenetics, and artificial intelligence.
Now, with Epi+Gen CHD, it looks specifically at the likelihood a patient or an individual is going to have a heart attack or die from a coronary heart disease event in the next three years. So it’s looking at events. And the product was developed exactly doing that. At baseline, none of these people had had an event. Within three years, they had developed an event. And so we built not just the algorithm to identify these individuals, but as a company, what we also do is develop proprietary assays, especially on the DNA methylation epigenetic side, to measure the biomarkers. So it’s a panel of eight biomarkers for EpiGen.
PrecisionCHD uses the same technology, but instead of eight biomarkers, it’s 16 biomarkers. They’re completely different biomarkers, again, genetic and epigenetic. And this time, what the algorithm is doing upon the evaluation of these biomarkers is to provide an idea whether we’re seeing a signal associated with the presence of coronary heart disease, meaning whether or not this person likely has, it’s a diagnostic test, whether they have coronary heart disease or not. What is truly transformative about our Epi+Gen test and PrecisionCHD test is that both of these tests are coupled to our proprietary platform called Actionable Clinical Intelligence.
Now, if we take a step back, I shared that we measure genetic and epigenetic markers. But we know that providers don’t treat genetic or epigenetic biomarkers. So we almost need a basis to decipher that data to say, “What do these markers mean? Where do they map to? What is driving this individual’s patient’s risk?” Because we’re measuring individual markers. And so ACI is almost that decoder. It takes these biomarkers for each individual patient, and it maps onto which of these markers are driving that patient’s signal, whether it’s the diagnostic signal or whether it’s the risk signal to say, “Where is that signal coming from?”
So it could come from inflammation, it could come from diabetes, it could come from cholesterol. But the main idea there is, “Are we thinking in the right direction? Are we thinking about personalizing the care for this patient?” And also, let’s be real. In clinical care, it’s always important for us to not just have hypothesis but also test our assumptions. We may think this person has elevated cholesterol, so as long as we treat cholesterol, we’re treating heart disease. But is that really true? We know that as a complex disease, heart disease is more than just one risk factor. So this ACI platform provides more insights into the personalized treatment.
And the other thing I would add is we’ve seen epigenetic markers change in as little as 90 days with intervention. And so ACI provides a way to longitudinally understand the changes in these biomarkers and maybe what interventions are working and maybe what’s not working as well as we thought it should.
Evgeniy Galimov
It’s very interesting. So basically, you’re saying that it’s possible to understand what are the drivers of a particular disease, in case, coronary heart disease?
Meesha Dogan
Yeah. And I think it’ll make intuitive sense to you because epigenetics is a reflection of our lifestyle and environment, which heart disease is only about 20% genetics. The rest of it is non-genetic in nature, which largely comes from our lifestyle and environment. And because epigenetics captures that and it’s dynamic, we’re able to map especially the epigenetic markers to factors that providers can think about treating and patients can think about changing in a way that could then potentially meaningfully change their risk and the management of coronary heart disease.
Evgeniy Galimov
But what kind of actionable parameters you can see, for example, for coronary heart disease?
Meesha Dogan
Yeah, that’s a great question. The ones that we specifically… I’ll use our PrecisionCHD test as an example. So that test has a total of 16 biomarkers. Six of it is epigenetic markers. The markers map to inflammation, diabetes, cholesterol, and specific pathways. And that again, is critical because these are the parameters clinicians are thinking about when it comes to treating these patients and for them to continue to monitor these patients as well.
Evgeniy Galimov
But compared to the conventional tools like Framingham score or pooled equations scores, how does your tool perform? And can you tell a little bit about that, and is it easier to explain the score compared to these tools?
Meesha Dogan
So when we’re thinking about the Framingham Risk Score and the ASCVD pool cohort equation, it’s closer to our Epi+Gen CHD test because Epi+Gen falls in the realm of more primary prevention risk assessment, which is where the Framingham Risk Score and ASCVD pool cohort equation risk calculators fall under. And one of the things we did was when we did the study, we did an initial study. The product was developed in the Framingham heart study cohort. It was validated in partnership with Intermountain Healthcare. Again, peer-reviewed, published. And in that peer-reviewed published manuscript, one of the things we outline is does our tests, the Epi+Gen CHD test, have more information content or better information content in which we’re identifying people better of the likelihood of having an event? And so that is sensitivity.
And when we looked at sensitivity of the Framingham Risk Score and ASCVD pool cohort equation versus our Epi+Gen CHD test, we found that Epi+Gen CHD was about 2.4 times more sensitive for women and about 1.7 times more sensitive for men when it comes to the ability to predict compared to the average between Framingham and ASCVD.
Now, if we think about why that is the case, and just from an intuitive nature, all of us involved in science, it will probably make sense that when we’re looking at heart disease, we’re really looking at three layers. We see the molecular layer, the intermediate proxy factors, the ones that Framingham Risk Score and ASCVD pool cohort equation measures, and then there’s heart disease.
What we’re doing with the Framingham Risk Score and ASCVD pool cohort equation, we’re making an a priori decision that these handful of markers are sufficient. As I shared, those markers are necessary. It’s necessary to know cholesterol, it’s necessary to treat cholesterol. It’s necessary to know hemoglobin A1C, it’s necessary to treat hemoglobin A1C. But they are just not sufficient because they’re not capturing so many more factors and element that we’ve shown that you could get from molecular markers, which in a way is agnostic because it changes for me versus you as opposed to we’re looking under the same lens for you and me with respect to cholesterol, and saying these cutoff points are what we should use to measure and what we should use to treat for all of us. When we were known that’s not necessarily the case, because if you think about it, two people who have the same cholesterol level do not have the same level of risk for a heart attack.
Evgeniy Galimov
And basically, so you share that your tool performs actually better in terms of sensitivity. But in terms of specificity, slightly less. I understand that it’s quite difficult to find the right balance between sensitivity and specificity, but obviously the lower specificity brings some problems with more money spent on studying patients who got the positive test. So what is the right balance and how do we address this?
Meesha Dogan
That’s a wonderful question. So basically to your point, sensitivity, anyone who’s done any level of machine learning knows that when you’re tuning a model, you have some give and take between sensitivity and specificity. When you’re trying to optimize for one, you’re giving a little with respect to the other. Now, let’s talk about, I’d like to turn your question on its head. And what I mean by that is if I failed to identify someone who was going to have an event, meaning I had a false negative, the cost of a false negative is someone’s going to go on to have a heart attack. And we know that for people who have coronary heart disease, sometimes the first symptom that they have, coronary heart disease, is a heart attack. And a lot of times a fatal heart attack. Meaning you’re not just having a heart attack, you could die from it. And downstream quality of life. And again, we’re talking about longevity, your health span versus your lifespan, right? Those things are impacted.
And that’s one of the arguments to maximize sensitivity, because no one wants to have a heart attack. But to your point, when you have a lower specificity, you could potentially have false positives. These are people who are not at elevated risk, but we’re flagging them as being at more elevated risk and they’re going on to get a secondary test.
Season 2, Episode 1: Dr. Irina Conboy and Serafina Mei
Season 2, Episode 2: Dr. Ming Yu and Namita Hattangady
Season 2, Episode 3: Dr. Jon Berner
Season 2, Episode 4: Dr. Meesha Dogan
Now, the question there is what is that secondary test? How much is that secondary test? And how invasive is that secondary test? And a secondary test could be as easy as a coronary calcium scan, which is really not that high of a burden. It’s for all intents and purposes, I think for some people it costs them about $60 or so to do it. And we actually took that exercise and did an economic model to understand, given our sensitivity and specificity, knowing that we were aiming more towards sensitivity, and again, balancing that out with specificity, there is actually data showing $42,000 or so in savings per quality adjusted life year and improve survival.
So I think the question you asked is very valid, but then the burden then falls to be able to go through and do the calculation to say, given these parameters compared to what is being done, is there still a cost savings? Because you’re spending more treating someone who has a heart attack for the rest of their life versus you’re paying that extra, call it about $60 or so for a coronary calcium scan. And that’s the exercise we undertook.
Evgeniy Galimov
Thank you. Great answer. I also would like to ask you about performance of the algorithm on different risk groups. Is it the same? Is it slightly different? Can you tell anything about calibration?
Meesha Dogan
Yeah, calibration as you’re probably really well aware, comes from data in the sense that it comes from representative data. And when we talk about representative data, what we’re talking about is first and foremost is the definition of what we’re doing. So if we’re saying with Epi+Gen CHD, it’s identifying those at risk of having a heart attack, is that data truly representative of people who went on to have a heart attack? So that’s one aspect. The other aspect is, is it the right population? And the right population could be, is the data set 80% 20 year olds? No 20-year olds really are having heart attacks, right? It’s not as representative. Also, are we looking at sex? Like male versus female? And the other aspect is we know that as a company that does include genetics in part, even though genetics is the smaller portion of what drives heart disease or at least coronary heart disease, there are genetic heart conditions. Are we including genetic markers that are not rare variants? Are they representative genetic markers that you see across different ethnicities?
And we checked all of those boxes when we were doing that. And more importantly, we checked those boxes across a study design that calibrated, to your point, the test in a test cohort, which is the Framingham heart study cohort. Part of it was testing and finalizing the model. Part of it that was not used in model development was used as a test data, but again, we know it’s not truly independent because it’s a proportion of the same cohort. But we went on to then try to understand the generalizability. So a complete external validation, completely independent cohort with Intermountain Healthcare.
And we did the same exercise for our precision CHD tests. But instead of just looking at Framingham heart study in Intermountain, we went on to do an additional independent validation in a whole other cohort as well called the University of Iowa Hospitals and Clinics cohort. So in a way, we have put together the separation of church and saint between development and validation and also thinking through the elements that go into that calibration in the form of data and also in the way we handle that data.
Evgeniy Galimov
Well, you mentioned that as I understood from papers, most of the data were for people with European ancestry. What will be your strategy to check or develop to better for more diverse population? Here in the UK, for example, we have quite diverse population, not only white and Black people, but also many Asians. What would be your approach in this case?
Meesha Dogan
Yeah. We had a couple of, not a large percentage, but a handful of people that were not European or African-American in our studies, just a handful of people. In those handful of people, we didn’t see a difference in the performance. And again, it would make intuitive sense because the genetic markers specifically that are being used as part of our tests are representative. And again, it has been shown in tens of thousands, hundreds of thousands of people of all different ethnic backgrounds has been shown to be representative. So that in itself is one of the main reasons we weren’t expecting to see a difference. And again, we did not in the people that we did.
But to your point, what we continue to intend to do, we’re a company that is very data-driven. If you remember, one of the first of the four things I shared was data. And it’s absolutely critical for companies like us to continue to do studies, clinical studies, research studies that are continuing to validate what we’re putting out there in multiple different areas, including, to your point, in more ethnically diverse cohorts. So it’s something that we always have ongoing and we’re very fortunate to be able to work with a lot of renowned providers and organizations, and we continue to do that, and we will continue to do that because it is the right thing to do and it only makes our products better over time.
Evgeniy Galimov
Also, curious, so is there any overlap between patients your algorithm predicted as having coronary heart disease and… Sorry, not coronary heart disease, high risk for cardiac event, and patients predicted by Framingham score, for example? Because I guess many people would be interested to understand are these people the same or not?
Meesha Dogan
No, that’s a great question. So the Framingham Risk Score, and I believe the ASCVD pool cohort equation as well, don’t quote me on that, but in part, they did use the Framingham heart study, which is the cohort that we used for our development as well. Now, I think the other way to answer what you just asked me was when we had the Framingham heart study data to develop the Epi+Gen CHD test, we also had data for these individuals to run the Framingham Risk Score and ASCVD pool cohort equation. So we actually ran both on the same individuals from the Framingham heart study data, which again, it’s a very well-known cohort, but we actually also did run it in the Intermountain cohort. So we did do head to head across two independent cohorts.
Evgeniy Galimov
So you see some overlap there or…
Meesha Dogan
That was the basis for the 2.4 times and 1.7 times more. Because we compared the output from the risk calculators versus the output from the Epi+Gen CHD test.
Evgeniy Galimov
So it looks like mostly, so your cohort of patients includes Framingham cohort of patients, but also predict more than that, right?
Meesha Dogan
Correct. Yes, that’s exactly it. So there was an overlap in individuals that we were able to do our test or our analysis on, and we were able to do the risk calculators on, and actually that’s the paper we published showing the overlap and head-to-head comparison in a way where we were able to show for the same individuals if we took all of their risk factors and put in the risk calculators versus we took their biomarkers and ran Epi+Gen CHD tests for the same individuals, are we able to predict them better using ours or using the risk calculators?
Evgeniy Galimov
Great. And obviously let’s just move more implementation side. How does this product make prevention more accessible and personalized?
Meesha Dogan
That I think is an innovation that came to us of its time. And the reason I say that is we were actually getting ready to launch EpiGen during the pandemic. Of course that was not our plan, but that’s how the timing came about. So initially when we were thinking about launching EpiGen, we were thinking about it a more straightforward blood test where the providers would have the ability to use the phlebotomist, their nurse, themselves, someone to draw a tube of blood and send it in. But what we realized during the time of COVID is we’re never going to practice medicine exactly the same way we did or even very close to the way we did from a healthcare delivery standpoint post-COVID as we did pre-COVID. And what I mean by that is if what we’ve seen trend-wise from post-COVID is we’ve seen a lot of medicine happening outside of the four walls of a clinic or a hospital.
And so for us, we realized that we needed to get with the times, we needed to get ahead of this and make sure our product is like a novel technology, but it has to meet the healthcare delivery where it’s going. And in that sense, telemedicine was another tool or solution that we were starting to use more and more. So we actually built our test delivery to be as modular as possible.
So there’s really two things that happens before we see a patient sample. There is the prescription that has to happen by a provider, and then there’s a sample collection that has to happen. On the prescription side, to make it modular and have multiple options, we coupled our tests to telemedicine. So remote option for prescription. We work with provider organizations where they order it when a patient comes into their office. And so that gives the flexibility of the mode to order the test, and patients can complete their health history questionnaires in person, they can complete it online, maybe at night when they’re sitting after work in their pajamas and filling out their information.
And then the second element is getting their blood drawn. Now on the provider side, a lot of our providers today do opt for tubes to keep in their office. So sample collection kits to keep in their office for them to draw the blood when their patients come in. Which is great, right? Your patient come in, a test is necessary, you order, you draw the blood, you’re done. To make a remote option available, we also have an at-home sample collection kit. So it uses the same Becton Dickinson tube, but instead of a vacutainer uses a microtainer that someone can collect at home. We also can send phlebotomists to someone’s house. And we actually do work with employers as well for the test for their employees.
And in that sense, we also do set up mobile health clinics. So a lot of times employers are thinking about utilization. More of their employees to utilize how can a technology meet their employees where they are? And we set up shop at their HQ. Because again, we’re not taking a big machine, we’re not radiating anyone. We’re literally showing up, setting up a bunch of phlebotomists continuously drawing blood as the test is ordered for these individuals.
So we’ve built those into different modules that we plug and play depending on who wants what. An employer could say, “We want telemedicine plus onsite. We want telemedicine plus a phlebotomist going to people’s house.” Or a provider can say, “I want to order the test for my patients, but I want you to send a kit for them to collect it at home.” We can think about all of those options. And that leads to accessibility and scalability because even if you think about rural America, which a new study came out not too long ago in the Journal of American College of Cardiology that said about 46% of counties in the US don’t have a cardiologist. What do you do? You cannot not prevent heart disease. It’s 80% preventable. If you’re not preventing it, then there’s something wrong with the infrastructure and the delivery of care. But with technology, and even technology like ours where behind the curtains is a complex technology but upfront implementation, it’s really simple. So there’s no excuse not to do prevention.
Evgeniy Galimov
Completely agree. And talking about personalization, so I found some publications where you say that it reflects smoking status. Can you tell us about any other modifiable factor or something which your test could reflect?
Meesha Dogan
Absolutely. As I shared earlier, the markers map to different pathways and different modifiable factors. Smoking is one of it, as I shared. Inflammation is another. Cholesterol is another. Specifically oxidation of cholesterol, which we know is very important in plaque buildup for instance. Diabetes as well. We have markers that looks at like HDL, particle size, atherosclerosis. So it certainly maps to multiple of these factors.
The study that you probably saw with respect to smoking cessation, we wanted to demonstrate how soon can you see an effect of an intervention in using epigenetic markers? Because we know epigenetic markers are dynamic, if you know what it’s mapping to, that’s step one. Step two is there needs to be an action. And step three is how do we measure how well that action worked? So in this case, we had markers mapping to smoking. The action was smoking cessation. The third thing now is to measure how quickly do we see an effect of that intervention in individuals? And that’s what we showed that you can see a significant change. I mean, you do see a change in like 30 and 60 days, but a significant enough change in about 90 days.
Evgeniy Galimov
And related question to this, what is the biggest change you saw so far in your patients in terms of improvement?
Meesha Dogan
That’s a great question. And I say that’s a great question because different providers practice medicine differently. I don’t think that’s a surprise to anyone. And the reason I preface with that is because I may think that I need to give my patients one year before I retest them. Some providers might say, “No, if I see a change in as little as 90 days, I want to see you in 90 days.” And mostly you can think of that in concierge practices or practices where you’re incentivized to spend more time with your patients. I mean, unfortunate aspects of healthcare. Some providers might say, “No, guidelines usually say you retest someone in two to three years when it comes to prevention.” So we’re still on, we have some of the data, like I shared, the 90 days is kind of where we’ve seen, again, we’ve saw changes in 30 days. We saw changes in 90 days. The most pronounced change we saw in… Sorry, 30 days, 60 days, and then the more pronounced change in 90 days.
And I’m excited for us to continue to have that data, especially from real world, because we know that in real world, behavioral change is really hard. Making these changes, whether it’s quitting smoking, losing weight, better diet, reducing stress to reduce inflammation, these things are personalized for the individual, but the individual and in context with the tools they have and their provider need to make those interactions in a way where there is behavioral change to see an effect. So again, I’m excited for us to see that over time and over various interventions. But again, for right now, 90 days is kind of the largest change we’ve seen.
Evgeniy Galimov
Great. Thank you. Just also curious about the adoption of your tools. First question is, why do you think doctors stick to the current conventional tools like Framingham score if they perform not that well? And second question is if your tool would be a black box, well, I know that your SNPs and methylation sites are linked to particular pathways, cardiovascular pathways, but if it would be black box, do you think it would be an obstacle for adoption?
Meesha Dogan
The answer to the first part of the question is similar to what I shared about the patients, behavioral change is hard. And for clinicians, it’s really not that different. Although I’ll share something. When we went to the American College of Cardiology, which is one of the flagship meetings we go to last year in 2023, I would say the better part of eight clinicians who came by our booth didn’t know what epigenetics was.
So first and foremost, the education aspect has to be there, which is something we do a lot. We do a lot of lunch and learns, we do educational sessions, we do ongoing training for providers to even adopt our test just because they have new providers coming in, they want to refresh. They want to know exactly how a report would be interpreted. And that comes over time, and we understand that. But the good news though is that is changing. This year in 2024, when we went to American College of Cardiology, I would say it was closer to about four clinicians out of 10 who didn’t quite know what epigenetics is.
So we’re seeing a shift, and I think AI has a part to play. Because maybe they don’t listen to epigenetics before they want to listen to AI, and then in part, they’re incorporating epigenetics. For us, it’s getting in the door, having the conversation with each provider in a way that makes sense for them, educating them. And I mean, despite that, late last year, you asked about adoption, late last year, we actually got a breakthrough… It’s called Breakthrough Innovation Contract, something along those lines from Vizient, which is the largest group purchasing organization in the country, which is the largest group purchasing organization in the country, which is the largest group purchasing organization in the country. And the reason for that is they didn’t have another test like ours on contract. And so they had to create a new path to be able to go down and award us a contract, which is nice. When you’re the first to do something, it’s great, but then you have other hurdles like education to work through.
And then your next question about black box, if trying get providers excited about things they can actually read and learn and point to, which are, again, to your point, not black box, is already a challenge, I don’t even want to think what it would take. And I just don’t think any technology when you are especially talking about clinical care, should be a black box. Because there is no way for you to use feedback or insights if you don’t know what could cause something to change or what could be affecting what you have in your black box. I don’t think it does anyone a service, whether it’s a provider, it’s a patient, the company who’s developing these technologies. Yeah, no. I hope no one does that. And I even think that would be much harder to get implemented and adopted.
Evgeniy Galimov
Thank you. And the last question about this as this area is, what do you think about health inequality and your tools? Is it possible to address health inequalities somehow?
Meesha Dogan
Yeah, that’s a great question. One of the key aspects about health inequality is accessibility. And I’m not saying accessibility alone is going to solve health inequality, but if you can’t get something to someone, or certain tools are only accessible because someone lives in a certain zip code or a certain part of the country or in urban areas versus rural areas. Conscious design around those technologies like ours needs to happen. Not because you’re going to solve health inequality, but you would help, one, make a difference. And I mean, I think that’s one of the aspects I’m really proud in the way we think about the delivery, as I told you. You don’t have a cardiologist in your county? That’s fine. You can still get access to the most cutting edge cardiology tests using epigenetics and AI. That helps address this in part some of it.
But I think the other aspect that is important is when companies like ours take those steps to say, we are going to do things differently because in part it’s going to help address health inequalities, others who are watching are going to want to do the same, and those who are paying and those who are evaluating technologies would start using that as a parameter to make those decisions.
So in a way, there are direct and indirect effects in addressing health inequality. And for us, one of the biggest way we’re looking to do that is first and foremost being targeted with our tools that are better, but also getting it in the hands of as many people as possible.
Evgeniy Galimov
Great, Meesha, thank you for answering about inequality. The next questions I would like to ask about relation to longevity area and methylation clocks. Do you know about correlation of your tool’s performance of your scores with any methylation aging clocks?
Meesha Dogan
Yeah, so that’s an interesting question. We’ve not done any direct analysis of any correlations or anything like that, but one of the hallmarks of the aging clocks, which there are multiple at this point. But I think most if not all of them, one of the key factors they point to is inflammation. And there was one specific clock, I want to say it might’ve been GrimAge, I could be wrong, but also talked about cardiovascular.
And I would say, again, this is my research side talking, I would be very surprised if the underlying factors that we are capturing molecularly, I would be surprised if it doesn’t overlap in a meaningful or in a relatively or pretty significant way with these clocks. And the reason I say that is because when we’re looking at the deviation between chronological agent biological age with these clocks, we’re seeing them as a result of people not just getting older, but them getting sicker. And with heart disease and also cancer being some of the leading sickness or conditions people get as they age, there should be an overlap, but I can’t say because we’ve not done any direct studies looking at them directly correlating with each other.
Evgeniy Galimov
Do you think your tool’s scores reflect more disease or aging of cardiovascular system?
Meesha Dogan
I would say it is going to be both. And the reason I say that is because if we’re looking at a disease like coronary heart disease, there are two really factors that are non-modifiable, meaning all of us, those two factors we can’t change, and that’s age and sex. So age in itself is a key factor in the risk of developing coronary heart disease. So as you grow older, your risk increases. And that’s the reason why I say that, yes, our tests, they’re picking up a specific definition of cardiovascular, in this case coronary heart disease. One is event. One is diagnostic. But it cannot be in the absence of the aging of the cardiovascular system because that is part of the stressors and the chronological age that feed into your risk of developing these events.
Granted, I can see people where their biological age may be much, much higher, like the delta, right between their biological age and their chronological age could be much bigger because they may have developed heart disease at a younger age, like more premature heart disease. So I really do think there would be the ability to pick up both. But again, I’ve not kind of split them up, so I can’t say to what extent it would pick up the cardiovascular aging, even though for reasons that I’ve said I believe it would.
Evgeniy Galimov
And how do you see implementation of your products in the longevity focused lifestyle?
Meesha Dogan
Yeah, the longevity folks are absolutely fascinating. I love the longevity movement. And the reason I share that is because the longevity movement has been able to bring together science and technology in a way that I’ve not seen many other movements do. And I call it a movement, not because it’s like a cult or something like that, but it has really gotten the attention of people to have conversations around their health, but also to pay attention to technologies and tools that help you track, measure and understand where you are. I know earlier I shared about healthspan and lifespan, I think they’ve really shed a light on those two things, which are important, how long you’re living and how much of you living is a healthy high quality life.
So yeah, we actually have very fascinating conversations with longevity providers. We work with longevity providers, and of course the lens in which they see their patients is slightly different than the more traditional medicine. But one of the trends I’ve actually also seen is providers who’ve had more traditional practices are now also launching longevity practices. Not because they can’t rope it all into one practice, but as I shared, you have to see things in a different lens when you’re looking at longevity and the kinds of questions people who come to you to get tested or be cared for from a longevity standpoint is different. So in a way we’ve seen them to be the more innovative providers.
Evgeniy Galimov
And you basically, do you see that they adopt your technology faster than traditional clinics, or not really?
Meesha Dogan
Yes, only because they’re not bogged down by clinical guidelines. They understand that the longevity movement has in a way questioned the rigid way. Some providers practice medicine, and rightfully or wrongfully so. You could have an argument either way. But faster is relative. They do dig into the technology more. I’ve noticed that they dig into a lot of primary literature more and they tend to have more questions on the technology side, including AI, what is your AI doing, how it was developed, those kinds of questions that maybe we don’t see from a more traditional provider who provides clinical care, maybe not as much in the longevity space. So they’ve been some of our bigger champions and they’re great to work with because they tell you exactly what they like, they tell you exactly what they don’t like, they give you so many ideas for improving your products. I have some great relationships with some of them, and they’re wonderful, wonderful providers.
Evgeniy Galimov
It’s great to hear that you like longevity-focused clinics.
Meesha Dogan
Absolutely, yes.
Evgeniy Galimov
And next question I would like to ask is what’s your vision for the future AI-based clinical decision tools? What do you see in this area?
Meesha Dogan
I know that AI and maybe precision are sometimes two overused words, but someone like me who comes, I’m an engineer by training. That’s what I did, my PhD and I’m biomedical engineer, prior to that I did my chemical engineering bachelor’s and master’s. And having invented the technology behind cardio and built a company around a novel technology, the meaning of artificial intelligence to someone like me in its use cases and the meaning of what precision medicine means for someone like me is profound in a way where I think the advancements in the science and technology and tools, our ability to measure better, our ability to gain insights at more granular level, our ability to move away from one size fits all decision-making, our ability to think about healthcare beyond just clinicians, including other stakeholders, like in our case, employers, they’re very innovative. They’re thinking about the best of the best benefits.
These things all will come together in a way where, you know how I shared COVID took us away from the four walls of relying on the four walls of a hospital or clinic and has taken medicine a lot outside of these four walls. I think these aspects of advancements is truly going to make medicine unrecognizable. And more than that, I think people are getting more informed and they’re going to demand it. They’re absolutely going to demand it. And the clinical side of things is going to have to evolve to meet that demand and meet what is possible because science and innovation is really not waiting for anyone. We’re going to move at a rapid pace and we just have to continue to find clinical champions, there are many of them who are advocating for their peers to move in that direction and move sooner rather than later.
Evgeniy Galimov
Great. Great vision. Well, the last question is can you share some examples of how your product has helped patients to manage cardiovascular risk?
Meesha Dogan
Yeah. I have so many fascinating stories. It’s one of the things I love doing when I’m checking in with my team or talking to the providers that we work with, is because I get to get a firsthand understanding of how has it really impacted patients. One example that I always love sharing is this person who, the provider had used our PrecisionCHD test to want to understand how well her management of coronary heart disease is going. Mainly looking at her epigenetic markers in understanding, do I have to see you as often as I’m seeing you? I think she was seeing her provider once a month just because she was someone that they were keeping a track of very, very closely. And based on what they were able to see with her molecular markers and how well they’ve been not changing, but how well they have been responding to interventions, they were able to change our cadence of seeing the provider and they had an understanding that we’re on the right track. Because my understanding was they tried a bunch of different things to make sure that they got the right formula to help her continue to manage.
And I find that to be… Again, we talk about things like utilization. How can we think about not just the right kinds of tests and treatments and management for patients, but also maybe what’s the right cadence to be seeing our patients based on some of these changes that we see? So that will always be one of my favorite examples.
Evgeniy Galimov
Thank you. Thank you very much. It’s very interesting to talk to you Meesha and all the best to your company and hope to see your products available in the UK too.
Meesha Dogan
Wonderful. Thank you so much for having me. This is such a fascinating conversation.
Evgeniy Galimov
Thank you.
Meesha Dogan
Wonderful. Take care.
