Nov 7, 2022
Dr. Carolyn Lam:
Welcome to Circulation on the Run, your weekly podcast summary and backstage pass to the Journal and its editors. We're your cohosts. I'm Dr. Carolyn Lam, Associate Editor from the National Heart Center, and Duke National University of Singapore. And...
Dr. Peder Myhre:
I'm Dr. Peder Myhre from Akershus University Hospital, and University of Oslo in Norway.
Dr. Carolyn Lam:
Peder, I'm so excited about our future discussion. It's about a very important topic of detecting atrial fibrillation in the population using wearable devices. It talks about the Fitbit Heart Study. So exciting, but we're going to keep the audience waiting a bit, because we're going to talk about some other things in the issue. And I would love to start with this now.
We know that fulminant myocarditis presentation is a rare and severe presentation of myocarditis. But, what is its natural history, and clinical features associated with poor outcomes? Peder, what do you think?
Dr. Peder Myhre:
Oh, that's a great question. We really don't know, because prior studies have been relatively small and selected. So Carolyn, let me know.
Dr. Carolyn Lam:
You're absolutely right. But today's paper from Professor Saito, from Nara Medical University in Japan and colleagues, is the largest nationwide cohort study of patients with histologically proven fulminant myocarditis presentation. They study 344 patients, hospitalized with histologically proven myocarditis, who underwent catecholamine and/or mechanical support from 235 cardiovascular training hospitals across Japan, between 2012 and 2017, and here's what they found.
Over a median follow up of 600 days, the accumulative risk of death or heart transplantation at 90 days was 29%. So, really high. These were the risk factors associated with a higher risk of death or heart transplantation, and they were non-sinus rhythm, older age, ventricular tachyarrhythmia, lower left ventricular ejection fraction. Severe histological damage was also associated with a worse 90 day outcome in lymphocytic myocarditis. Cool, huh?
Dr. Peder Myhre:
Oh wow. That was some really solid data. And now Carolyn, I'm going to take us over to the world of preclinical science. And the next paper entitled at “APIC Associated De Novo Purine Synthesis is Critically Involved in Proliferative Arterial Disease” by Yuqing Huo from Augusta University in Georgia.
Dr. Carolyn Lam:
Cool.
Dr. Peder Myhre:
And as you know, Carolyn, vascular smooth muscle cells are extremely important in vascular health. They're located in the medial layers of arteries, and normally exhibit a contractile phenotype that contributes to the regulation of blood vessel tone, blood flow distribution, and blood pressure in normal mature blood vessels. And in response to disease processes, the vascular smooth muscle cells are switched to an activated synthetic and proliferative phenotype, that contribute to the development of a variety of arterial diseases, including atherosclerosis, in-stent restenosis, and bypass graft occlusion. And nucleotides that we are familiar with, such as ATP and GTP, are essential for a large number of biological processes in cells, including proliferation.
And Carolyn, the previous studies have demonstrated that de novo synthesis of purine is a critical pathway for nucleotide synthesis. And in this study, the authors assessed the role of de novo synthesis of purine in vascular smooth muscle cells by using knockout mice.
Dr. Carolyn Lam:
Oh, that was beautifully explained. Thanks, Peder. So what did they find?
Dr. Peder Myhre:
So the authors found that the de novo purine synthesis was increased in proliferative vascular smooth muscle cells. Moreover, they identified an important enzyme in the process called A-P-I-C, APIC. Which was observed in the neointima of the injured vessels, and atherosclerotic lesions in both mice and humans.
Finally, they showed that in a mouse model with knocked out APIC, the atherosclerosis and arterial restenosis was attenuated.
Dr. Carolyn Lam:
Cool. So tell us what the clinical implications are.
Dr. Peder Myhre:
So these findings provide novel insights into the reprogramming of purine metabolism underlying vascular smooth muscle cells proliferation in the development of arterial disease. And that targeting APIC may be a promising therapeutic approach to combat arterial diseases.
So Carolyn, please tell me about your next paper.
Dr. Carolyn Lam:
Ah, thanks, Peder. Well, back to the clinical world, this time, talking about arrhythmogenic right ventricular cardiomyopathy. We know that is characterized by progressive cardiomyocyte loss and fibro fatty replacement. And we know that patients with this a ARVC are at risk for life-threatening ventricular arrhythmias and sudden cardiac death.
The placement of an ICD is a crucial component of ARVC management. But arrhythmic risk stratification and the selection of the optimal candidates for ICD, especially for primary prevention of sudden cardiac death, has, of course, been challenging.
As background, a ventricular arrhythmia risk calculator, in patients without previous sustained ventricular arrhythmias, has been proposed, and includes seven clinical variables derived from non-invasive tests that are routinely performed in these patients. However, the possibility of integrating additional parameters, such as ventricular tachycardia inducibility on programmed ventricular stimulation, with this risk calculator, has been suggested, but not conclusively investigated in a large cohort. And so, here comes corresponding author, Dr. Cadrin-Tourigny, from Montreal Heart Institute and colleagues, who studied 288 patients with a definite ARVC diagnosis, no history of ventricular arrhythmias at diagnosis, and programmed ventricular stimulation performed at baseline. And these patients were identified from six international ARVC registries.
Dr. Peder Myhre:
Oh wow. So we're talking risk stratification for patients with ARVC. Such an interesting topic, Carolyn. So please tell me, what did they find?
Dr. Carolyn Lam:
So, programmed ventricular stimulation significantly improved risk stratification, above and beyond the calculator predicted risk of ventricular arrhythmias, in a primary prevention cohort of patients with ARVC. And this was mainly for patients considered to be at low and intermediate risk by the clinical risk calculator. If negative, its high negative predictive value of 93% in low and intermediate risk patients, may support the decision to forego ICD use in some patients. So, programmed ventricular stimulation results may be applied to the non-invasive ARVC risk calculator, in a two step approach to facilitate personalized decision making for ICD in such patients.
Dr. Peder Myhre:
Thank you, Carolyn. That was a great summary and a great paper. So we're going to move in to see what else is in the mail bag, Carolyn.
Dr. Carolyn Lam:
You bet. There's a letter by Dr. Agirbasli regarding the article, “Coronary Artery and Cardiac Disease in Patients with Type Two Myocardial Infarction, A Prospective Cohort Study,” and this, followed by a response by Dr. Chapman.
There's an ECG Challenge by Dr. [Jingnan] Han, entitled, “Tachycardia Associated with Pacing.”
From our own Molly Robbins, we have highlights from the Circulation Family of Journals. And she covers the experience with stereotactic radio ablation and electrical storm, reported in Circulation: Arrhythmia and Electrophysiology.
The impact of accessibility to primary care on hypertension awareness and control is reported in Circulation: CV Quality and Outcomes.
There's an analysis of lifestyle factors and their impact on the risk of heart failure by background genetic risk, and that's in Circulation: Heart Failure.
There's a deep learning model of PET scans and coronary flow reserve reported in Circulation: CV Imaging.
And finally, OCT based measurement of stent expansion and associations with outcomes are presented in Circulation: CV Interventions.
A lot.
Dr. Peder Myhre:
Yeah, and there's more, Carolyn. In this issue, there is an extensive Frontiers review by the AF-SCREEN International Collaboration, entitled, “Consumer LED Screening for Atrial Fibrillation.”
There is also a Research Letter by corresponding author Qi Fu, from University of Texas Southwestern Medical Center entitled, “Neuro Cardiovascular Dysregulation During Orthostasis in Women with Posttraumatic Stress Disorder.”
And finally, a Research Letter by Pankaj Arora from University of Alabama entitled, “Mechanical Circulatory Support Devices Among Patients with Familial Dilated Cardiomyopathy, Insights from the INTERMACS.”
Dr. Carolyn Lam:
That's awesome, Peder. Thank you. Now let's go onto our feature discussion on atrial fibrillation detection and the Fitbit Heart Study, shall we?
Today's feature discussion is about the Fitbit Heart Study, and none other than the first and corresponding author Dr. Steven Lubitz, from Massachusetts General Hospital in Boston to join us today. Steve, welcome. Congratulations. Am I right to say, this is the largest study of its kind to look at the detection of atrial fibrillation using wearable devices?
Dr. Steven Lubitz:
Thanks for having me, Carolyn. And that's right, this is.
Dr. Carolyn Lam:
Oh my gosh. Okay. Tell us all about it, what you did, what you found.
Dr. Steven Lubitz:
Well, thanks, Carolyn. So as we know, undiagnosed atrial fibrillation is a potential hazard that can cause strokes. And if we can identify people who have undiagnosed atrial fibrillation early, we may be able to prevent strokes. In addition, undiagnosed atrial fibrillation may be associated with additional morbidity, which can be addressed through a number of different ways, if we can detect atrial fibrillation. Obviously, the challenge is to detect atrial fibrillation.
We also know that people are increasingly wearing devices that have sensors on them, specifically using photoplethysmography technology, which can detect the pulse rate. Software algorithms can now be developed, that can assess that pulse rate for regularity or irregularity. But they really need to be assessed and validated, to minimize the potential for false positives, which can have obviously, downstream adverse consequences of their own, if atrial fibrillation is incorrectly identified or diagnosed as a result.
As I was mentioning, we developed this novel software algorithm with frequent overlapping photoplethysmography, post tachogram sampling, which is unique. And then we tested the algorithm's positive predictive value for undiagnosed AFib in a large scale remote clinical trial, using a range of Fitbit wearable fitness trackers and smart watches.
It was a remote trial, so participants were invited. These were people who already had a Fitbit account, they were invited to participate. And in span of just a few months, in the middle of the pandemic, over 455,000 people signed up to participate in the study. And so, big thank you to all of the participants in the study.
Dr. Carolyn Lam:
Wow, that is big. And what did you find?
Dr. Steven Lubitz:
So of the 455, over 455,000 participants that enrolled, over 4,000, had an irregular heart rhythm detection and received a notification. And after inviting those participants to attend a telehealth visit, and at that telehealth visit, the telehealth provider confirmed eligibility criteria, confirmed that they didn't have preexisting atrial fibrillation, for example, and a variety of other inclusion/exclusion criteria.
They were mailed a one week ECG patch, that they applied themselves, and then returned that ECG patch. So in the end, after those exclusions, in participants that returned analyzable patches, 1057 participants were included in this ECG monitoring analytic cohort, of whom, 340 had atrial fibrillation during that ECG patch monitoring period.
The primary endpoint of the study was the positive predictive value of irregular heart rhythm detection that occurred during the ECG patch monitoring period. So a participant had to have an irregular heart rhythm detection to get notified that they were eligible to meet with a telehealth provider and receive an ECG patch monitor. And then, they had to have another irregular heart rhythm detection during ECG patch monitor wear. So the primary outcome was the positive predictive value of the first irregular heart rhythm detection for concurrent atrial fibrillation that occurred during ECG patch monitoring.
Dr. Carolyn Lam:
Okay. Cool. So many questions here, but maybe you should tell us the results first.
Dr. Steven Lubitz:
Sure. So the primary endpoint, the positive predictive value of the IHRD during ECG patch monitoring was 98.2% in the overall cohort. And it was similar between men and women, and those aged 65 or older, or those aged less than 65. And I should mention that, in this study, about 13% of participants enrolled in this study overall, were above the age of 65.
Dr. Carolyn Lam:
And you included more women than in prior similar studies. Right, Steve?
Dr. Steven Lubitz:
Yeah.
Dr. Carolyn Lam:
I was going to congratulate you for that.
Dr. Steven Lubitz:
Yeah, that's right. That's right. We're very excited to see that.
Dr. Carolyn Lam:
Okay, so that's cool. Wow. A positive predictive value of 92%. So couple of things here with-
Dr. Steven Lubitz:
98.
Dr. Carolyn Lam:
Sorry, 98%. That's right. Wow. Okay. Now with this AFib detection, it's always about duration. Right? And what do you call a positive alert? Could you maybe elaborate a bit about that here?
Dr. Steven Lubitz:
Sure. So I think this is an important point. A few points. One, the algorithm is designed. This particular algorithm requires at least 30 minutes of an irregular pulse to be detected, in order for a detection to occur. Which means that, this is unlikely to be detecting trivial amounts of atrial fibrillation. And indeed, that's what we observed. We observed that the median burden of atrial fibrillation was 7% among those who had AFib on the ECG patch monitor. We observed that the median longest episode of atrial fibrillation was seven hours. And just by way of comparison, in other studies in which ECG patch monitors have been distributed to people without this irregular pulse pre-screening, the burden is usually on the order of only a couple of percent, tops. So this, by nature, these types of algorithms, and this algorithm specifically, probably enriches for individuals who have a higher burden of atrial fibrillation. Meaning that, if these detections occur, then it's probably not detecting trivial amounts of atrial fibrillation.
Dr. Carolyn Lam:
Right. And a lot of it seems to send a very clear message that this study, and perhaps even the algorithm, is designed to be specific. Right? So that duration, as well as what you used as the outcome. How much price do you pay in terms of sensitivity? Do you know what I mean? Since we optimized for specificity, am I right to say that?
Dr. Steven Lubitz:
Sure, that's a great point. The algorithm is really optimized for specificity, as you mentioned. And although we didn't specifically calculate the sensitivity of the algorithm, in a secondary analysis, we examined the sensitivity of an IHRD during that ECG patch monitoring period, to detect any AFib that was documented on the ECG patch monitor, and it was about 67%.
So we know that we probably don't detect some atrial fibrillation. Largely, that's a function of this technology at the moment. It's very difficult to assess the pulse rate during periods of activity in motion. So a lot of these algorithms, and this algorithm in particular, doesn't operate during periods of motion. The accelerometers and the devices can tell the algorithm that motion is occurring, and then the algorithm won't operate on that information at that time. So a lot of this has to do with limitations of the technology at the moment.
Dr. Carolyn Lam:
Ah. So the detection probably occurs best at rest or at night.
Dr. Steven Lubitz:
That's exactly right. And we encourage participants to wear their devices at nighttime during the study.
Dr. Carolyn Lam:
Oh, cool. And then of course, I suppose a question you'd anticipate, I mean, we know about the Apple Heart Study, we know about the watch study, and how does this compare? How is this technology different, and the results?
Dr. Steven Lubitz:
Essentially, one of the most remarkable things about these studies is that, it appears that this pulse rhythm pre-screening really enriches substantially for people who have atrial fibrillation. So for example, in the Fitbit Heart Study, we observed that about 32% of people who had an irregular heart rhythm detection and then returned an ECG patch monitor, had AFib on it. And by comparison, in the Apple Heart Study, that number was about exactly the same, just over 30% or so.
So when we further compare this pre-screening type approach to confirming atrial fibrillation, using an ECG patch monitor, with other approaches in which say, elderly individuals were mailed ECG patch monitors to screen for atrial fibrillation, we usually only see detection in the order of four to 5% of people. So this irregular pulse based pre-screening markedly enriches for atrial fibrillation. And we also know, this is only a one week ECG patch monitor, and if we monitor people longer than one week, we're likely to detect more atrial fibrillation, since this is often paroxysmal atrial fibrillation that we're detecting.
So there are a lot of similarities, and I think the point is that, these types of consumer electronic devices are going to be great tools for identifying undiagnosed atrial fibrillation in the community. I think we have a lot of challenges ahead of us, in terms of figuring out how to integrate that information into our routine healthcare workflow, and counseling consumers and users of these types of technology on exactly what they should be doing when they do get an alert. And then also, counseling providers on how to act on these findings, what they mean and how accurate the technology is.
Dr. Carolyn Lam:
Yeah. And I appreciated a sentence in your manuscript that talks of, what are our society guidelines going to say? If you could look into a crystal ball now, Steve, based on what you found, what would you advise both patients and clinicians, if you don't mind?
Dr. Steven Lubitz:
Well, I think that, in short, if a clinician is alerted by a patient, that they received in a regular heart rhythm detection on their device, in short, I would say, don't blow it off. Take it seriously. Because the odds are, that it does represent an abnormality, and the odds are that that abnormality is atrial fibrillation. And given the potential adverse consequences of undiagnosed atrial fibrillation, there's a real opportunity to intervene, and prevent morbidity in the patient. And then, if you're a consumer who happens to have one of these devices, and you've turned on this feature, and hopefully you have, if you do have an alert, don't blow it off. Contact a provider. Because it may very well mean that you have an irregular heart rhythm that merits attention, and could be addressed to prevent downstream consequences and morbidity for you.
Dr. Carolyn Lam:
Nice. And keep your Fitbit on at night.
Dr. Steven Lubitz:
Yes. And if you do want to maximize the utility of these algorithms that use photoplethysmography, probably wearing them at nighttime will maximize the sensitivity, or utility of the devices and algorithms.
Dr. Carolyn Lam:
Aw, that's just great. What nice take home messages. Thank you so much, Steve, for publishing this really unique and important study in Circulation.
So audience, you heard it right here on Circulation on the Run. From me, Greg, and Peder, please do tune in again next week.
Speaker 4:
This program is copyright of the American Heart Association 2022. 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, please visit ahajournals.org.