March 2023 Discover CircRes

This month on Episode 46 of Discover CircRes, host Cynthia St. Hilaire highlights four original research articles featured in the March 3 and March 17th issues of Circulation Research. This episode also features an interview with Dr Andrew Hughes and Dr Jessilyn Dunn about their review, Wearable Devices in Cardiovascular Medicine.   Article highlights:   Delgobo, et al. Deep Phenotyping Heart-Specific Tregs   Sun, et al. Inhibition of Fap Promotes Cardiac Repair After MI   Sun, et al. Endosomal PI3Kγ Regulates Hypoxia Sensing   Johnson, et al. Hypoxemia Induces Minimal Cardiomyocyte Division   Cindy St. Hilaire:        Hi, and welcome to Discover CircRes, the podcast of the American Heart Association's Journal, Circulation Research. I'm your host, Dr Cindy St. Hilaire from the Vascular Medicine Institute at the University of Pittsburgh, and today I'm going to share four articles selected from the March 3rd and March 17th issues of CircRes. I'm also going to have a discussion with Dr Andrew Hughes and Dr Jessilyn Dunn about their review, Wearable Devices in Cardiovascular Medicine. And the Review is also featured in our March 3rd issue.   Cindy St. Hilaire:        First, the highlights. The first article I'm going to present is Myocardial Milieu Favors Local Differentiation of Regulatory T-Cells. The first author is Murilo Delgobo and the corresponding author is Gustavo Campos Ramos. After myocardial infarction, the release of autoantigens from the damaged heart cells activates local and infiltrating immune cells such as the T-cell. Studies in mice have shown that fragments of the muscle protein myosin can act as autoantigens, and these myosin fragments are the dominant driver of the T-cell response.   But how do these myosin specific T-cells behave in the damaged heart to drive inflammation and repair is unknown. To find out, Delgobo and colleagues studied endogenous myosin specific T-cells, as well as those transferred into recipient mice. They found, whether exogenously supplied or endogenously created, the myosin specific T-cells that accumulated in the animals' infarcted hearts tended to adopt an immunosuppressive T-regulatory phenotype.   Strikingly, even if the exogenous cells were differentiated into inflammatory TH-17 cells prior to transfer, a significant proportion of them were still reprogrammed into T-regs within the heart. Although cells pre-differentiated into an inflammatory TH-17 phenotype were less inclined to change after the transfer, the results nevertheless indicate that, by and large, the infarcted heart promotes T-cell reprogramming to quell inflammation and drive repair. Yet exactly how the heart does this is a question for future studies.   Cindy St. Hilaire:        The next article I'm going to present is titled Inhibition of FAP Promotes Cardiac Repair by Stabilizing BNP. The first authors of the study are Yuxi Sun and Mengqiu Ma, and the corresponding author is Rui Yue, and they are from Tongji University. After myocardial infarction, there needs to be a balance of recovery processes to protect the tissue. Fibrosis, for example, acts like an immediate bandaid to hold the damaged heart muscle together, but fibrosis can limit contractile function.   Similarly, angiogenesis and sufficient revascularization is required to promote survival of cardiomyocytes within the ischemic tissue and protect heart function. To better understand the balance between fibrotic and angiogenic responses, Sun and colleagues examined the role of fibroblasts activated protein, or FAP, which is dramatically upregulated in damaged hearts, and brain natriuretic peptide, or BNP, which promotes angiogenesis in the heart.   In this study, they found that genetic deletion or pharmacological inhibition of FAP in mice reduces cardiac fibrosis and improves angiogenesis and heart function after MI. Such benefits are not seen if BNP or its receptor, NRP-1, are lacking. The in vitro experiments revealed that FAP's protease activity degrades BNP, thus inhibiting the latter's angiogenic activity. Interestingly, while FAP is upregulated in the heart, its levels drop in the blood, showing that BNP inhibition is localized. Together, these results suggest that blocking FAP's activity in the heart after MI could be a possible strategy for protecting the muscle's function.   Cindy St. Hilaire:        The next article I want to present is Hypoxia Sensing of Beta-Adrenergic Receptor is Regulated by Endosomal PI-3 Kinase Gamma. The first author of this study is Yu Sun, and the corresponding author is Sathyamangla Naga Prasad. Hypoxia is the most proximate acute stress encountered by the heart during an ischemic event. Hypoxia triggers dysfunction of the beta-adrenergic receptors, beta-1AR and beta-2AR, which are critical regulators of cardiac function.   Under normoxic conditions, activation of PI3K-gamma by beta-adrenergic receptors leads to feedback regulation of the receptor by hindering its dephosphorylation through inhibition of protein phosphatase 2A or PP2A. Although it is known that ischemia reduces beta-adrenergic receptor function, the impact of hypoxia on interfering with this PI3K feedback loop was unknown.   Using in vitro and in vivo techniques, this group found that activation of PI3K-gamma underlies hypoxia sensing mechanisms in the heart. Exposing PI3K-gamma knockout mice to acute hypoxia resulted in preserved cardiac function and reduced beta-adrenergic receptor phosphorylation. And this was due to a normalized beta-2AR associated PP2A activity, thus uncovering a unique role for PI3K-gamma in hypoxia sensing and cardiac function.   Similarly, challenging wild-type mice post hypoxia with dobutamine resulted in an impaired cardiac response that was normalized in the PI3K-gamma knockout mice. These data suggests that preserving beta-adrenergic resensitization by targeting the PI3K-gamma pathway would maintain beta-adrenergic signaling and cardiac function, thereby permitting the heart to meet the metabolic demands of the body following ischemia.   Cindy St. Hilaire:        The last article I want to highlight is Systemic Hypoxia Induces Cardiomyocyte Hypertrophy and Right Ventricle Specific Induction of Proliferation. First author of this study is Jaslyn Johnson, and the corresponding author is Steven Houser, and they're at Temple University.   The cardiac hypoxia created by myocardial infarction leads to the death of the heart tissue, including the cardiomyocytes. While some procedures such as reperfusion therapy prevent some cardiomyocyte death, true repair of the infarcted heart requires that dead cells be replaced. There have been many studies that have attempted new approaches to repopulate the heart with new myocytes. However, these approaches have had only marginal success.   A recent study suggested that systemic hypoxemia in adult male mice could induce cardiac monocytes to proliferate. Building on this observation, Johnson and colleagues wanted to identify the mechanisms that induced adult cardiomyocyte cell cycle reentry and wanted to determine whether this hypoxemia could also induce cardiomyocyte proliferation in female mice.   Mice were kept in hypoxic conditions for two weeks, and using methods to trace cell proliferation in-vivo, the group found that hypoxia induced cardiac hypertrophy in both the left ventricle and the right ventricle in the myocytes of the left ventricle and of the right ventricle. However, the left ventricle monocytes lengthened while the RV monocytes widened and lengthened.   Hypoxia induced an increase in the number of right ventricular cardiomyocytes, but did not affect left ventricular monocyte proliferation in male or in female mice. RNA sequencing showed upregulation of cell cycle genes which promote the G1 to S phase transition in hypoxic mice, as well as a downregulation of cullen genes, which are the scaffold proteins related to the ubiquitin ligase complexes. There was significant proliferation of non monocytes in mild cardiac fibrosis in the hypoxic mice that did not disrupt cardiac function.   Male and female mice exhibited similar gene expression patterns following hypoxia. Thus, systemic hypoxia induced a global hypertrophic stress response that was associated with increased RV proliferation, while LV monocytes did not show increased proliferation. These results confirm previous reports that hypoxia can induce cardiomyocyte cell cycle activity in-vivo, and also show that this hypoxia induced proliferation also occurs in the female mice.   Cindy St. Hilaire:        With me today for our interview, I have Dr Andrew Hughes and Dr Jessilyn Dunn, and they're from Vanderbilt University Medical Center. And they're here to discuss the review article that they helped co-author called Wearable Devices in Cardiovascular Medicine. And just as a side note, the corresponding author, Evan Brittain, unfortunately just wasn't able to join us due to clinical service, but they're going to help dissect and discuss this Review with us. Thank you both so much for joining me today. Andy, can you just tell us a little bit about yourself?   Andy Hughes:             Yeah, thank you, Cindy. I'm Andy Hughes. I'm a third year medicine resident at Vanderbilt University who is currently on an NIH supported research year this year. And then will be applying to cardiology fellowships coming up in the upcoming cycle.   Cindy St. Hilaire:        Great, thank you. And Jessilyn, I said you are from Vanderbilt. I know you're from Duke. It was Evan and Andy at Vanderbilt. Jessilyn, tell us about yourself.   Jessilyn Dunn:             Thanks. I am an Assistant Professor at Duke. I have a joint appointment between biomedical engineering and biostatistics and bioinformatics. The work that my lab does is mainly centered on digital health technologies in developing what we call digital biomarkers, using data from often consumer wearables to try to detect early signs of health abnormalities and ultimately try to develop interventions.   Cindy St. Hilaire:        Thank you. We're talking about wearable devices today, and obviously the first thing I think most of us think about are the watch-like ones, the ones you wear on your wrists. But there's really a whole lot more out there. It's not just Apple Watches and Fitbits and the like. Can you just give us a quick summary of all these different types of devices and how they're classified?   Jessilyn Dunn:             Yeah, absolutely. We have a wide variety of different sensors that can be useful. A lot of times, we like to think about them in terms of the types of properties that they measure. So mechanical properties like movement, electrical properties like electrical activity of the heart. We have optical sensors. And so, a lot of the common consumer wearables that we think about contain these different types of sensors.   A good example that we can think about is your consumer smartwatch, like an Apple Watch or a Fitbit or a Garmin device where it has something called an accelerometer that can measure movement. And oftentimes, that gets converted into step counts. And then it may also have an optical sensor that can be used to measure heart rate in a particular method called PPG, or photoplethysmography. And then some of the newer devices also have the ability to take an ECG, so you can actually measure electrical activity as well as the optical based PPG heart rate measurement. These are some of the simpler components that make up the more complex devices that we call wearables.   Cindy St. Hilaire:        And how accurate are the measurements? You did mention three of the companies, and I know there's probably even more, and there's also the clinical grade at-home ECG machines versus the one in the smartwatch. How accurate are the measurements between companies? And we also hear recent stories about somebody's Apple Watch calling 911 because they think they're dead, things like that. Obviously, there's proprietary information involved, but how accurate are these devices and how accurate are they between each other?   Jessilyn Dunn:             This is a really interesting question and we've done quite a bit of work in my lab on this very topic, all the way from what does it mean for something to be accurate? Because we might say, "Well, the more accurate, the better," but then we can start to think about, "Well, how accurate do we need something to be in order to make a clinical decision based off of that?" And if it costs significantly more to make a device super, super accurate, but we don't need it to be that accurate to make useful decisions, then it actually might not be serving people well to try to get it to that extreme level of accuracy.   So there are a lot of trade-offs, and I think that's a tough thing to think about in the circumstances, is these trade-offs between the accuracy and, I don't know, the generalizability or being able to apply this to a lot of people. That being said, it also depends on the circumstances of use. When we think about something like step counts, for example, if you're off by a hundred step counts and you're just trying to get a general view of your step counts, it's not that much of a problem.   But if we're talking about trying to detect an irregular heart rhythm, it can be very bad to either miss something that's abnormal or to call something abnormal that's not and have people worried. We've been working with the Digital Medicine Society to develop this framework that we call V3, which is verification, analytic validation and clinical validation. And these are the different levels of analysis or evaluation that you can do on these devices to determine how fit for purpose are they.   Given the population we're trying to measure in and given what the goal of the measurement is, does the device do the job? And what's also interesting about this topic is that the FDA has been evolving how they think about these types of devices because there's, in the past, been this very clear distinction between wellness devices and medical devices. But the problem is that a lot of these devices blur that line. And so, I think we're going to see more changes in the way that the FDA is overseeing and potentially regulating things like this as well.   Cindy St. Hilaire:        These consumer-based devices have started early on as the step counters. When did they start to bridge into the medical sphere? When did that start to peak the interest of clinicians and researchers?   Jessilyn Dunn:             Yeah, sure. What's interesting is if we think back to accelerometers, these have been used prior to the existence of mobile phones. These really are mechanical sensors that could be used to count steps. And when we think about the smartwatch in the form that we most commonly think of today, probably looking back to about 2014 is when ... maybe between 2012, 2014 is when we saw these devices really hitting the market more ... Timing for when the devices that we know as our typical consumer smartwatch today was around 2012 to 2014.   And those were things that were counting steps and then the next generation of that added in the PPG or photoplethysmography sensor. That's that green light when we look on the back of our watch that measures heart rate. And so, thinking back to the early days, probably Jawbone, there was a watch called Basis, the Intel Basis watch. Well, it was Basis and then got acquired by Intel. Fitbit was also an early joining the market, but that was really the timing.   Cindy St. Hilaire:        How good are these devices at actually changing behavior? We know we're really good at tracking our steps now and maybe monitoring our heartbeat or our oxygen levels. How good are they at changing behavior though? Do we know yet?   Andy Hughes:             Yeah, that's a great question and certainly a significant area of ongoing research right now with physical activity interventions. Things that we've seen right now is that simple interventions that use the wearable devices alone may not be as effective as multifaceted interventions. And what I mean by that is interventions that use the smartwatch but may be coupled with another component, whether that is health education or counseling or more complex interventions that use gamification or just in time adaptive interventions.   And gamification really takes things to another level because that integrates components, competition or support or collaboration and really helps to build upon features of behaviors that we know have an increased likelihood of sustaining activity. With that being said, that is one of the challenges of physical activity interventions, is the sustainability of their improvements over the course of months to years.   And something that we have seen is the effects do typically decrease over time, but there is work on how do we integrate all of these features to develop interventions that can help to sustain the results more effectively. So we have seen some improvement, but finding ways to sustain the effects of physical activity is certainly an area of ongoing research.   Cindy St. Hilaire:        I know it's funny that even as adults we love getting those gold stars or the circle completions. All of these devices, whether it's smartwatches like we're just talking about, or the other things for cardiac rehabilitation, they're generating a ton of data. What is happening with all this data? Who's actually analyzing it? How is it stored and what's that flow through from getting from the patient's body to the room where their physician is looking at it?   Andy Hughes:             And that is certainly a challenge right now that is limiting the widespread adoption of these devices into routine clinical care is, as Jessilyn mentioned. The wearables generate a vast amount of data, and right now, we need to identify and develop a way as clinicians to sort through all of the noise in order to be able to identify the information that is clinically meaningful and worthy of action without significantly increasing the workload.   And a few of the barriers that will be necessary in order to reach that point is, one, finding ways to integrate the wearables' data into the electronic health record and also developing some machine learning algorithms or ways with which we can use the computational power of those technologies to be able to identify when there is meaningful data within all of the vast data that comes from wearables. So it's somewhere that certainly we need to get to for these devices to reach their full clinical potential, but we are limited right now by a few of those challenges.   Jessilyn Dunn:             I was just going to say, I will add on to what Andy was saying about this idea behind digital biomarkers because this fits really nicely with this idea that giving people this huge data deluge is not helpful, but if we had a single metric where we can say, "Here's the digital biomarker of step count, and if you're above some threshold, you're good to go. And if you're below some threshold, some intervention is needed." That's a lot of the work that we've been doing, is trying to develop what are these digital biomarkers and how can they be ingested in a really digestible way?   Cindy St. Hilaire:        Yeah, that's great. Regarding the clinical and the research grade devices, I know a Fitbit or Apple Watch can sometimes be used for those, but I guess I'm talking also about the other kind of more clinically oriented devices, how good is compliance and how trustable is that data? Everybody's on probably their best behavior when they're in the office with the physician or if they're on the treadmill in the cardiac lab, but home is a different story. And what don't we know about compliance when people are out of the office and the reliability of that data that's generated in that space?   Andy Hughes:             I think you touched on a really important point right here, and one of the potential advantages of these wearable devices is that they provide continuous long-term monitoring over the course of weeks to months to years as opposed to those erratic measurements that we get from the traditional office visits or hospitalizations where, for example, the measurements we're taking are either in a supervised environment with a six-minute walk distance, for example, or self-reported or questionnaires.   So we build upon that information, but then additionally, we go beyond the observer effect where many individuals, the first week or two that you're wearing this new device, you may be more prone to increase your activity because you know that you're being monitored or you have this novel technology, but as you wear it for months to years, you outgrow those potential biases and you really can garner more comprehensive information.   In terms of compliance, we can speak to some of the research studies that have either really struggled with compliance and that limits the interpretability of their results and something we'll need to address in the future, but I think that's something that can be addressed with future studies keeping in mind all of the advantages that these devices offer compared to some of the traditional measures that we have used in the past.   Cindy St. Hilaire:        With all this data we're collecting, whether it be biological data or even just behavioral data, have we actually learned anything new? And I mean that in terms of All Of Us study this, I don't know, it was like 5,000 patients I think, and lo and behold, it found out that higher step count correlated with lower risk for a ton of diseases, which is not exactly groundbreaking. So are we, at this point in time, learning anything new from the use of these at-home devices, or are they really just able to help us enforce what we thought we knew regarding behavior?   Andy Hughes:             I think these devices have certainly provided some novel insights that build upon our understanding of physical activity. Many of us can hypothesize that decreased activity would have poor outcomes on health, which the studies have demonstrated in many facets. But in reference to All Of Us study that you mentioned, I think it's interesting to look as well at some of the diagnoses or conditions that were associated with decreased activity.   For example, reflux disease was also highlighted in that study, which may not have been identified if we didn't have the vast data and ability to really look for associations with diseases that have not been previously studied or thought to be related to physical activity. So I think that's one of the strong features of that database, is the wealth of knowledge that really will be hypothesis generating and help to inform future studies as we look even beyond cardiovascular conditions.   Cindy St. Hilaire:        One question, and you did bring it up in a bit of the discussion in your piece, is the bias that is in these devices. We know from COVID at-home pulse oximeters do not work as efficiently on darker skin. We actually know that going into bathrooms with the hand sensors that spit out the paper towels. So what kind of disparities or biases do these devices create or reinforce in the population?   Jessilyn Dunn:             This is such a critical topic because a lot of these issues had been discovered retrospectively because the people who were developing the technologies were not the representative of the people who were using the technologies. I think that's something that across the board we've been looking at from device development to AI implementation, which is having people who are going to be using the devices in the process of developing the technology and having voices heard from across the board.   We did a detailed look when we were evaluating devices for their accuracy at this exact question of where the heart rate sensors in smartwatches use optical based technology. And there was some evidence that was also an issue for people with varying skin tones, for people with wrist tattoos or more hair or freckles. And so, we did a deep dive and the generation of devices that we looked at which would meet this study was probably about three years ago.   We didn't see any discrepancies. And so, that's just one study and there are many more to be done, but I think prior to the technology development as well as once the technology comes out, keeping an eye on how that technology is doing, whether there are continued reports of failure of the technologies is really important. And there are a lot of ways that we can be vigilant about that.   Cindy St. Hilaire:        Yeah, that's great. And so, Andy, regarding patient populations, I can also see perhaps socioeconomic implications of this because smartwatches are not cheap. So how do we see that in terms of helping our patients? Are we going to be able to get a smartwatch through our insurance company?   Andy Hughes:             I think that's one of the really important next steps, is finding ways to make sure that as we advance the field of wearable devices in clinical care, that we recognize some of the existing inequities in terms of access to care, access to digital technologies that currently exist, and find ways by partnering with health insurance companies and the industry and providers and members of that community, finding ways to not only advance wearables, but use it in a way that we can decrease health disparities by really helping to increase access for these digital technologies to the underserved communities.   Jessilyn Dunn:             Yeah, the beauty of these technologies is that truthfully, at their core, they're very cheap. They're not difficult to develop, they're not difficult to build and disseminate. So a lot of what we think about is the infrastructure that goes around these devices. Does it require a smartphone to transfer data? Does it require internet access? What are the other pieces that need to be in place for these devices to work within an ecosystem? So this starts to get to questions beyond the devices themselves, but there's certainly a lot to think about and be done in the area of equity and ensuring that these devices can help everyone.   Cindy St. Hilaire:        And there's also the, I guess, ethical considerations of who owns this data. Obviously, if it's a consumable that you went and bought at Target, that's probably different than the one you're getting from your cardiologist. But who owns the data? Who has access to it? And are there any cases in the literature where an individual who's had certain measurements taken, have those measurements come back to bite them?   And I guess I'm thinking of something like cardiac rehab. If a patient doesn't get up and move enough or doesn't follow their physical therapy enough or lose weight quick enough, could their insurance coverage get cut? Could their premiums go up? What safeguards are in place for these very tricky situations? Are there safeguards in place?   Andy Hughes:             And on the clinical side, I think it will be important to treat this information just like any other protected health information that we have as part of the electronic health record. And so, there will be inherently safeguards around that in a similar manner for how we treat other protected health information.   But I think another important component of that will be a very clear consent policy when we reach the point that patients are consenting to include this information and their electronic health record, in terms of what the proposed benefits are and the potential risks associated with it, because it really is a vast amount of unique data that needs to be protected and safeguarded. And part of that comes by treating it as protected health information, but we will also need to make sure that there's a very clear consent policy that goes with it.   Cindy St. Hilaire:        Yeah. What do we see as the next steps in wearable devices? What do you guys see as the next big thing? I know one's coming from the actual AI and device side of things, and the other one is coming from the clinical side of things. What do each of you see as the next thing in this field?   Jessilyn Dunn:             I think on the device and AI side of things, I think we're thinking toward improving battery life, increasing the suite of sensors that are being added to these devices so we have a wider variety of measurements that are more representative of physiology, and then better algorithms to have better detection of sleep or activity or certain types of activity or certain types of arrhythmias. This combination of hardware and software and algorithms, I think coming together as all of these different pieces evolve will show us some really cool technology in the years to come.   Andy Hughes:             And I think from a clinical side, it's really twofold moving forward. I think as Jessilyn mentioned, there's a lot of novel sensor technologies that have a lot of exciting and evolving potential that we can hopefully integrate into the clinical space, but on the other hand, it's how can we use these wearable devices to enhance traditional therapies that we're already using?   For example, if we take the heart failure population, is there a way that we can use the wearable devices and the existing measurements with heart rate and physical activity and blood pressure to find a way to improve remote management and safely up-titrate guideline directed medical therapy, which are medications that we know have clinical benefit. But can we augment their clinical benefit and their utility by using some of the existing technologies that we already have?   And then lastly, building upon the initial studies with larger trials in more diverse generalizable populations to really enhance our understanding of the benefits that these devices may have for different cardiovascular conditions.   Cindy St. Hilaire:        Well, this was wonderful. Dr Andrew Hughes and Dr Jessilyn Dunn, thank you so much for joining me. The review, Wearable Devices in Cardiovascular Medicine, will be out in our March 3rd issue of Circulation Research. I forget which one, so I'll have to edit that out. Thank you so much for joining us, and I learned a ton. This was great.   Jessilyn Dunn:             Thank you.   Andy Hughes:             Thank you.   Cindy St. Hilaire:        That's it for our highlights from the March 3rd and March 17th issues of Circulation Research. Thank you for listening. Please check out the Circulation Research Facebook page and follow us on Twitter and Instagram with the handle @CircRes and #DiscoverCircRes. Thank you to our guests, Dr Andrew Hughes and Dr Jessilyn Dunn.   This podcast is produced by Ishara Ratnayaka, edited by Melissa Stoner, and supported by the editorial team of Circulation Research. Some of the copy texts for the highlighted articles is provided by Ruth Williams. I'm your host, Dr Cindy St. Hilaire, and this is Discover CircRes, you're on-the-go Source for the most exciting discoveries in basic cardiovascular research.   This program is copyright of the American Heart Association, 2023. The opinions expressed by speakers in this podcast are their own, and not necessarily those of the editors or of the American Heart Association. For more information, visit ahajournals.org.