Jul 4, 2022
This week, please join author Fabian Eichelmann and Associate Editor Svati Shah as they discuss the article "Deep Lipidomics in Human Plasma: Cardiometabolic Disease Risk and Effect of Dietary Fat Modulation."
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 co-hosts. I'm Dr. Carolyn Lam, Associate Editor from the National Heart Center and Duke National University of Singapore.
Dr. Greg Hundley:
And I'm Dr. Greg Hundley, Associate Editor Director at the Pauley Heart Center at VCU Health in Richmond, Virginia. Well, Carolyn, this week's feature, we are going to get into the world of lipidomics and understand how some lipid metabolites may be more predictive of cardiovascular events above and beyond the conventional serum lipoproteins, like HDL and LDL. But before we get to that, how about we grab a cup of coffee and start with some of the other articles in the issue? Would you like to go first?
Dr. Carolyn Lam:
Sure thing. This first paper is about the basilar artery. Have you ever heard the analogy that the basilar artery is the neurologists' equivalent of our cardiologists' left main coronary artery? Well put, isn't it? Well, that's from the editorial that accompanies this paper, but basically, occlusion of either artery can be fatal without rapid re-perfusion. So that's why the basal artery is the neurologists' left main coronary artery. Re-perfusion therapies for acute basilar artery occlusion include thrombolysis or mechanical endovascular thrombectomy.
Dr. Carolyn Lam:
In today's issue, Professor Hu from University of Science and Technology of China, and Professor Nogueira from University of Pittsburgh School of Medicine in the US, these are the co-corresponding authors, and their colleagues, reported the outcome of the attention registry of more than 2000 patients with acute basilar artery occlusion who enrolled prospectively and consecutively at 48 sites in China from 2017 to 2021, and followed for the primary outcome of a favorable neurological functional outcome defined as a modified Rankin score of zero to three at 90 days.
Dr. Greg Hundley:
Wow, Carolyn, I love that analogy. So the basilar artery is kind of similar in the brain to the left main coronary artery in the heart. Whoa, what did they find here?
Dr. Carolyn Lam:
So, in this nationally representative observational study, the authors found a significant association between endovascular thrombectomy and better functional outcomes and survival at 90 days in patients with acute basilar artery occlusion, and this was compared to chemical thrombolysis. Now, notably, this relationship was modified by the baseline NIH stroke scale. Specifically, patients with baseline NIH stroke scale of 10 or more had an increased rate of favorable outcome when treated with endovascular thrombectomy, whereas no significant beneficial effect was seen in patients with baseline NIH stroke scores of less than 10. Now, all this is discussed in a beautiful editorial by Dr. Hankey, entitled, "Endovascular therapy for acute basilar artery occlusion."
Dr. Greg Hundley:
Oh, beautifully stated, Carolyn. What a great article. But guess what, Carolyn? I've got a quiz for you.
Dr. Carolyn Lam:
Uh-oh.
Dr. Greg Hundley:
This one's open answer, so it's not multiple choice. Can you name a unique feature of zebra fish pertinent to the study of cardiovascular disease?
Dr. Carolyn Lam:
Okay. Watch me hedge there because, first of all, I'm a daughter of a zoologist. So my dad would have a heart attack if I couldn't say something about the zebra fish. So, what I do know is the zebra fish is an excellent animal model for genetic studies of heart generation, basically development. So, there must be something really cool there about how we can observe that.
Dr. Greg Hundley:
Excellent, Carolyn. So very well done. As you know, and your dad knows, certain non-mammalian species like zebra fish, have an elevated capacity for innate heart regeneration. Now, understanding how heart regeneration occurs in these contexts can help illuminate cellular molecular events that can be targets for heart failure prevention or treatment, your area of expertise. The epicardium, the mesothelial tissue layer that encompasses the heart, is a dynamic structure that is essential for cardiac regeneration in zebra fish, and these authors, led by Dr. Jinhu Wang from Emory University performed single cell RNA sequencing and identified seven epicardial cell clusters in adult zebra fish, three of which displayed enhanced cell numbers during regeneration.
Dr. Carolyn Lam:
Oh, interesting, Greg. So did these cell clusters provide some clues that could be applied clinically?
Dr. Greg Hundley:
Yes. Carolyn. So these authors identified that these subsets of epicardial cells emerge in post embryonic, zebra fish and sponsor regions of active cardio myogenesis during cardiac growth and regeneration. And as the heart achieves its mature structure, these cells facilitate extracellular matrix hyaluronic acid deposition to support formation of the compact muscle layer of the ventricle. These cells associate with the function of the hyaluron and proteoglycan link protein 1 or HAPLN1 paralogue in production and organization of hyaluronic acid containing matrix in cardiac injury sites and thereby enable normal cardiomyocyte proliferation and muscle generation. And so Carolyn, potentially in the future targeting hyaluronic acid regulation by manipulation of HAPLN1 in human epicardial cells could potentially modulate cardiac repair after myocardial infarction.
Dr. Carolyn Lam:
Well, thanks, Greg. That was awesome. Well, the next paper I want to tell you about is one in which a novel ECG based machine learning approach was used to determine and predict multiple structural heart conditions. So the authors led by Dr. Chen from Department of Translational Data Science and Informatics at the Geisinger Health System in Danville, Pennsylvania. So these colleagues hypothesized that a composite model would yield higher prevalence and positive predictive value to facilitate meaningful recommendations for echocardiography.
Dr. Greg Hundley:
Oh wow, Carolyn. Machine learning, it's just emerging everywhere these days. So don't we need a large data set to do this?
Dr. Carolyn Lam:
Absolutely, and listen to how large this is. So using more than 2.2 million ECGs linked to electronic health records and echocardiography reports from almost 500,000 adults between 1984 and 2021, the authors trained machine learning models to predict the presence or absence of any of seven echo confirmed diseases within a year, and the composite model and the composite label that they used included moderate or severe valve disease and reduced ejection fraction. So their composite recommend model where reco is E-C-H-O we used age, sex ECG traces, and had an area under the receive operating curve of 0.91 and a positive predictive value of 42% at 90% sensitivity with a composite label prevalence of 17.9%. Whereas the individual disease models had area under curve ranging from 0.86 to 0.93 and lower positive predictive values from about 1% to 31%.
Dr. Carolyn Lam:
So in summary, they showed that an ECG based machine learning model using a composite endpoint can identify a high risk population for having undiagnosed clinically significant structural heart disease while outperforming the single disease models and improving practical utility with higher positive predictive values. So this approach may facilitate targeted screening with echo to improve under diagnosis of structural heart disease.
Dr. Greg Hundley:
Wow, Carolyn, really great article from the world of machine learning. Well, how about we jump to some of the other articles in the issue and I can go first. There are two Research Letters. The first Research Letter comes from Professor Adlam entitled "Pregnancy and Spontaneous Coronary Artery Dissection Lessons from Survivors and Nonsurvivors." And our own Dr. Joe Hill, our Editor-in-Chief also has a Research Letter entitled “Impaired AMP Kinase Signaling in HFpEF Associated Atrial Fibrillation.”
Dr. Carolyn Lam:
Wow, what an issue filled with great stuff. There's an AHA update by Dr. Churchwell on promoting nutrition security through policies and programs. And there are highlights from the circulation family of journals by our own Molly Klemarczyk, now known as Molly Robbins. I'd love to tell you a little bit about it. The association of new onset AF with cardiovascular outcomes in patients hospitalized with COVID-19 are described in Circ Arrhythmia, and EP. Rates of cardiovascular and cerebral vascular disease mortality among Asian subgroups are presented in Circ CV, Quality and Outcomes. Blood pressure and glycemic control are presented in patients with heart failure in Circ Heart failure. The associations of atrial update with technetium 99 pyrophosphate scans for transthyretin amyloid cardiomyopathy with incident atrial fibrillation and possibly earlier diagnosis of amyloid is presented in Circ Cardiovascular Imaging.
Dr. Carolyn Lam:
And finally the outcomes associated with larger burdens of residual thrombus after aspiration thrombectomy for STEMI are presented in Circ CV Interventions. Isn't that cool? And finally, we've got an On My Mind paper by Dr. Arany on “It’s Time to Offer Genetic Testing to Women with Peripartum Cardiopathy”. So that wraps it up, Greg. Let's go to our feature discussion, shall we?
Dr. Greg Hundley:
You bet.
Dr. Greg Hundley:
Welcome listeners to this feature discussion on July 5th, and we are very fortunate today. We have with us Dr. Fabian Eichelmann from the German Institute of Human Nutrition in Potsdam, Germany, and our own associate editor, Dr. Svati Shah from Duke University in Durham, North Carolina. Welcome to you both. Fabian, we're going to start with you. Can you describe for us some of the background information that went into the preparation of your study and what was the hypothesis that you wanted to address?
Dr. Fabian Eichelmann :
Yeah, sure. So first of all, thanks for the invitation to speak here. So this project was basically comes from a collaboration between us and a group in Redding, UK and we are part of a consortium called FAME, which is short for fatty acid metabolism. And there, we are interested in the health effects of fatty acid metabolism in general. And in this paper that we did, this was particularly cardiometabolic diseases. And I think this is no surprise that we look at lipid metabolism in this context, because there's so many really now also causal factors, lipoproteins, total triglycerides for specific cardiometabolic outcomes. So this is the reason why we wanted to look at it. And through this collaboration, we were also able to harness the potential from two different study designs that I probably will go into later, but which really gave us an opportunity to really generate I think, quite interesting insights.
Dr. Greg Hundley:
Very nice. And so what was that hypothesis?
Dr. Fabian Eichelmann:
So the hypothesis was that since the lipid metabolism has formally only been mostly in the clinic, at least been measured by lipoproteins and total triglycerides, for example, but the lipidome of plasma, for example, is really rich. It's really heterogeneous and it contains many different lipid classes and different fatty acids. And through novel technologies and in this case, lipidomics, you can really dive in really deeply and look at this in a specific manner. And then the idea was to really look at this and potentially identify lipids that would be associated or could surface as biomarkers for cardiometabolic diseases and at the same time, if those lipids were also sensitive to a dietary intervention that really tried to modulate the dietary intake of a fat.
Dr. Greg Hundley:
Very nice. And so how did you set this up? What was your study design and what was your study population?
Dr. Fabian Eichelmann:
So we had two different study populations for this. So the first one was the EPIC-Potsdam, which is a cohort study, a large scale cohort study here in Potsdam, which started in the nineties. And there, we basically associated baseline concentrations of these lipid measurements with later on occurring incident cardio metabolic outcomes. And in this case, this was type two diabetes and primary CVD and CVD in our case meant myocardial infarction and stroke. And we did that we checked which lipids would be statistically significantly associated after multiple testing with at least one of these outcomes. And then we took those lipids further into intervention trial, which is called the DIVAS trial in Redding, as I said, UK. And there, they had basically a dietary intervention trial that really wanted to assess if the change in the fatty acid proportions in the diet affects the lipids. So there we had the lipidomics measurements at baseline and after four months, and then we compared three different trials to each other.
Dr. Greg Hundley:
Very nice. And so how many patients did you include and who were these patients? Men and women? And did they have, for example, prior cardiovascular disease?
Dr. Fabian Eichelmann:
So in EPIC-Potsdam, that's a population based cohort study. All of the participants were drawn from the registries and invited. So these were apparently healthy people. And in those we did these association analyses and those we used the design, which is a case cohort design, which is a sub sample of the whole cohort, which is a really effective way and efficient way of analyzing biomarker projects. And there we had in total 1,262 control participants, and then later on additionally, a 775 type two diabetics and 551 CVD cases. In the DIVAS trial, that was a trial where participants also men and women, which was also the case in EPIC-Potsdam were invited, and they were at a higher risk of at the higher cardiovascular disease risk, which was measured by score, but they didn't have any prior cardiovascular diseases. And those were 113 participants that were randomized to one of these diets.
Dr. Greg Hundley:
Right. So it sounds like two studies. One, a large case control study and looking at different plasma lipid concentrations in two separate groups. And then the second was a randomized trial, a smaller trial of 113 individuals looking at a dietary intervention. So with that established, tell us your study results.
Dr. Fabian Eichelmann:
Yeah. So in the first step, as I said, where we associated lip concentrations to later occurring disease, we found from the 282 lipids that we looked at, 69 were really associated to at least one of these outcomes. And interesting here we saw that only eight were associated to both outcomes and 49 were specific to cardiovascular disease and 12 for specific to type two diabetes. And from those 69, we found 19 were also sensitive to the dietary intervention, and what was really striking here was that of these 19, 17 were perfectly in agreement with a suggested beneficial effect, meaning that those lipids that were associated in the EPIC-Potsdam studies on the cohort study with a higher disease risk were reduced by these diets or in the opposite direction, we saw those lipids that were associated with lower risk were increased by these diets. So this was quite a striking observation there.
Dr. Greg Hundley:
So it sounds like from the lipidomics analysis, there was a construct of certain blood lipid markers that were associated with cardiovascular events, and then in your randomized trial, you were able to modify those by different dietary interventions?
Dr. Fabian Eichelmann:
Exactly.
Dr. Greg Hundley:
So listeners, now we're going to turn to our own associate editor, Dr. Svati Shah. And Svati, you see many papers come across your desk. What attracted you to this particular paper and how do we put this study's results in the context with other studies that have been published in really the sphere of lipidomics research?
Dr. Svati Shah:
Yeah. Thank you, Greg. I just want to point out that this is a really elegant translational study. I think these papers can be very complicated to understand, and I think the authors did a fantastic job of really laying out how you can combine cutting edge what we call omics, using these cutting edge technologies, but applying them to human cohorts with a very strong clinical lens. So it's not just what do we learn about biology, but also what do we learn about biomarkers that might be relevant to how we take care of patients? And that's really one of the biggest things I loved about this study is sort of, you get to have your cake and eat it too. You get to learn about biology, but with a very strong clinical lens towards identifying clinically relevant biomarkers. I think another really important strength of this study, which differentiates it is this sort of use of really cutting edge lipidomics.
Dr. Svati Shah:
So this is a subset of omics where we're really looking at these granular lipid classes. And some of the clinicians might say, well, we measure cholesterol, why is this different? And you know, Greg, really what lipidomics allows us is a much more granular snapshot of these complex lipid species that are only grossly captured by the cholesterol levels that we would measure normally if we were seeing a patient in clinic. So to be able to get this really granular snapshot of what is happening to lipid biology and how it might relate to cardiovascular events, diabetes, I think is really important. And finally, I think the coolest part about this study is, in other studies we always have a little bit of a hard time with what we call confounders. And what does that mean? That means there may be other things for why you're seeing these biomarkers associated with your disease and those might be uncaptured things, things that you didn't measure in this study. And we call that residual confounding.
Dr. Svati Shah:
And I think the authors in this study, not only statistically adjusted for those potential confounders, but also importantly, the DIVAS study, where they took the biomarkers that they found from EPIC-Potsdam and said, do they change with the dietary intervention? And in fact, they did find that many of these lipid, these granular lipid species improved, meaning they went in the proper direction in terms of your health with a diet that was higher in unsaturated fats. So really proving not only the potential biology of the benefit of diets enriched in unsaturated fats, but also that these particular biomarkers are modifiable, so they're able to be changed even in this case within 16 weeks with just a dietary intervention. So to me that really was just a beautifully laid out study that highlights really what translational omics and these biomarker studies can do as we think about the clinical care of patients.
Dr. Greg Hundley:
Very nice. And so it sounds like listeners, moving beyond the lipoproteins, LDL, maybe total cholesterol and these granular lipid species cholesterol esters, free fatty acids, fingo-lipids, glycerophospholipid, et cetera, that's what we were studying in this particular manuscript. So, well, let's turn back to Fabian. And Fabian, what do you see is the next study to really be performed in this sphere of research?
Dr. Fabian Eichelmann:
I think what would be really interesting and what would really kind of prove what we saw is if you could find a way that an intervention, be it like a drug intervention and not diet because we looked at dietary intervention, that kind of shows the same as we saw, but also that really specifically only alters these lipids and how obviously not really feasible, but if that would be going on for a long enough period, if you also saw these effects that we now saw after four months would also affect the health outcomes instead of just only these proxies.
Dr. Greg Hundley:
Very nice. And Svati, how about you? What do you see as the next study that might be informative in this sphere of research?
Dr. Svati Shah:
Yeah. Great question, Greg. I mean, to me, I really think about this gap that we have in actually translating these findings in how we take care of patients. So again, really provocative results here. We have really significant P values, strong effect sizes, biomarkers that are modifiable, but in the paper they show that it adds on top of a clinical model, so what we might use as clinicians in the clinic that these biomarkers may help on top of what we already know about patients. But we really need to implement these findings and study that implementation for how this might in a real world setting actually change outcomes in patients and how we can actually help explain to clinicians how these results might be beneficial for clinical care.
Dr. Svati Shah:
So on top of what Fabian already said, I think really implementation science is a huge gap in how we take these translational omics discoveries and use them in support of improving patient care. We have lots to learn about these lipid biomarkers and lots more discovery science that can be done. As Fabian said, can we find drugs that might beneficially modified these lipid subspecies? But again, I think this gap in implementation science is really important.
Dr. Greg Hundley:
Very nice. Well listeners, we want to thank Dr. Fabian Eichelmann from the German Institute of Human Nutrition in Potsdam, Germany, and our own associate editor, Dr. Svati Shah from Duke University in Durham, North Carolina for bringing us these really provocative results, highlighting the identification of several lipids and their association with cardiometabolic disease risk. And then also nested within the same paper, a subset of individuals undergoing a randomized clinical trial demonstrating benefit by a dietary fat intervention, and there possibly supporting the substitution of dietary saturated fatty acids with unsaturated fatty acids perhaps as a potential tool for primary disease prevention.
Dr. Greg Hundley:
Well, on behalf of Carolyn and myself, we want to wish you a great week and we will catch you next week on the run.
Dr. Greg Hundley:
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.