Circulation April 11, 2023 Issue

This week, please join author Kavita Sharma and Associate Editor Svati Shah as they discuss the article "Myocardial Metabolomics of Human Heart Failure With Preserved Ejection Fraction."

Dr. Greg Hundley:

Welcome listeners, to this April 11th issue of Circulation on the Run. And I am one of your cohosts, Dr. Greg Hundley, director of the Pauley Heart Center at VCU Health in Richmond, Virginia.

Dr. Peder Myhre:

And I am Dr. Peder Myhre from Akershus University Hospital, and the University of Oslo in Norway.

Dr. Greg Hundley:

Well, Peder, wow. This week's feature discussion, very interesting. We spend a lot of time, especially with our colleague, Dr. Carolyn Lam, on heart failure preserved ejection fraction. But this week's feature discussion, it's going to focus on some of the myocardial metabolomics in this condition. But before we get to that, how about we grab a cup of coffee, and jump into some of the other articles in the issue? How about if I go first?

Dr. Peder Myhre:

Let's go, Greg.

Dr. Greg Hundley:

Okay. So Peder, some believe that cardiovascular disease may be the main reason for stagnant growth in life expectancy in the United States since 2010. And so, the American Heart Association, as you know, recently released an updated algorithm for evaluating cardiovascular health. Life's Essential 8, and it has a very nice score. So these authors, led by Dr. Lu Qi, from Tulane University, aimed to quantify the associations of the Life Essential 8 scores with life expectancy in a nationally representative sample of US adults. And the team included 23,000 non-pregnant non- institutionalized participants who were age 20 to 79 years, who participated in the National Health and Nutrition Examination survey, or NHANES, from 2005 to 2018. And whose mortality was identified through linkage to the National Death Index, from the period extending through December of 2019.

Dr. Peder Myhre:

Oh wow. So really, a validation of the Life's Essential 8. Greg, that's so interesting. What did they find?

Dr. Greg Hundley:

Right Peder, as you say, very interesting. So here are some of the data, and let's itemize them. So, during a median of 7.8 years of follow up, 1,359 total deaths occurred. Now, the estimated life expectancy at age 50 was 27.3 years, 32.9 years, and 36.2 years, in participants with low Life's Essential 8 scores, less than 50. Moderate, so Life's Essential 8 scores of greater than or equal to 50, but less than 80. And then, high scores, greater than 80. Okay? So equivalently, participants with high Life's Essential 8 scores had an average of 8.9 more years of life expectancy at age 50, compared to those with low scores.

Next, on average, 42.6% of the gained life expectancy at age 50, from adhering to sort of that cardiovascular health, those recommendations, was attributable to reduced cardiovascular death.

Next, significant associations with the Life's Essential 8 score and life expectancy were observed in both men and women.

Next, similarly significant associations of cardiovascular health, Life's Essential 8, with life expectancy were observed in non-Hispanic Whites and non-Hispanic Blacks, but not in those originating from the country of Mexico.

So Peder, finally, in summarizing all of this, adhering to the cardiovascular health lifestyle, defined by the Life's Essential 8 score, it was related to a considerably increased life expectancy. However, because of the findings from the individuals from the country of Mexico, more research is needed to be done in some of these minority groups, and particularly, those of Hispanic ethnicity, and perhaps other races.

Dr. Peder Myhre:

Oh, wow. Very interesting. And I would love to learn more about this subgroup analysis in future studies.

So Greg, the next paper is about the hospitalization for heart failure measures. Because contemporary measures of hospital performance for heart failure hospitalization, the 30-day risk standardized readmission and mortality rate, are estimated using the same risk adjusted model and overall event rate for all patients. Thus, these measures are mainly driven by the care quality and outcomes for the majority racial ethnic groups, and may not adequately represent the hospital performance for patients of Black or other races. And in this study, led by co-corresponding authors, Mentias from Cleveland Clinic and Pandey from University of Texas Southwestern Medical Center, the authors used fee for service Medicare beneficiaries from 2014 to 2019 hospitalized with heart failure, in hospital level 30 day risk standardized remission and mortality rates were estimated using traditional race agnostic models and the race specific approach, with measures derived separately for each race ethnicity group.

Dr. Greg Hundley:

Ah, very interesting, Peder. So what did they find from this study?

Dr. Peder Myhre:

So the study included more than 1.9 million patients, comprising of 75% White patients, 15% Black patients, and 10% patients of other races, with heart failure from 1,860 hospitals. And compared with the race agnostic model, composite race-specific metrics for all patients demonstrated stronger correlation with 30 days readmissions. And that is correlation coefficient 0.78 versus 0.63, and 30 day mortality rate 0.52 versus 0.29 for Black patients. In concordance in hospital performance was for all patients and patients of Black race was also higher with race specific as compared to race agnostic metrics.

So Greg, the authors conclude that among patients hospitalized with heart failure race specific 30 day risk standardized remission and mortality rates are more equitable in representing hospital performance for patients of Black and other races.

Dr. Greg Hundley:

Very nice, Peder. What a beautiful summary in a very elegant study. Peder, myocardial insulin resistance is a hallmark of diabetic cardiac injury. However, the underlining molecular mechanisms for this relationship remain unclear. Now, recent studies demonstrate, that the diabetic heart is resistant to several cardioprotective interventions, including adiponectin and pre-conditioning. The universal quote, unquote, resistance to multiple therapeutic interventions suggest, impairment of the requisite molecule, or molecules, involved in broad pro survival signaling cascades. Now caveolin is a scaffolding protein coordinating trans-membrane signaling transduction. However, the role of caveolin-3 in diabetic impairment of cardiac protective signaling and diabetic ischemic heart failure is unknown. And so these investigators, led by Dr. Xinliang Ma, from Thomas Jefferson University, studied mice fed a normal diet or high fat diet for two to 12 weeks, and subjected them to myocardial ischemia and reperfusion.

Dr. Peder Myhre:

Oh wow. What an interesting preclinical science paper, Greg. What did they find?

Dr. Greg Hundley:

Right. So the authors found that nitration of caveolin-3 at tyrosine 73 and resulted signal complex dissociation was responsible for cardiac insulin adiponectin resistance in the pre-diabetic heart. And this contributed to ischemic heart failure progression. Now, early interventions preserving caveolin-3 centered signal zone integrity was found to be an effective novel strategy against diabetic exacerbation of ischemic heart failure. And Peder, I think these very exciting results suggest that this is a new area of research and further experiments are warranted.

And there's a very nice editorial by Professor Heidenreich, entitled “Pursuing Equity in Performance Measurement.

Well Peder, there's some other articles in this issue, and we'll dip in this week to the mail bag, for a Research Letter from Professor Hibbert, entitled “Utility of a Smartphone Application in Assessing Palmar Circulation Prior to Radial Artery Harvesting for Coronary Artery Bypass Grafting.”

Dr. Peder Myhre:

That is so cool. And we also have a Letter from Dr. Kim, regarding the article entitled, “Detection of Atrial Fibrillation in a Large Population Using Wearable Devices: The Fitbit Heart Study.”

Dr. Greg Hundley:

Very nice. Well, how about we get along to one of Carolyn's favorite topics, heart failure with preserved ejection fraction, and learn more about myocardial metabolomics?

Dr. Peder Myhre:

Can't wait.

Dr. Carolyn Lam:

Today's feature discussion is on my favorite topic, heart failure with preserved ejection fraction, or HFpEF. But today, what we're focusing on is truly novel. We are looking at the myocardial metabolomics of human HFpEF, very, very valuable data and insights. We're so pleased to have with us the corresponding author of today's feature paper, Dr. Kavita Sharma, who's from the Johns Hopkins University School of Medicine, and our associate editor, Dr. Svati Shah, who's, of course, from Duke University School of Medicine.

So welcome Kavita and Svati. Kavita, if I could start by, please put us and bring us all to the same level of knowledge, by perhaps explaining in simple terms, what is metabolomics? And what is normal versus perhaps abnormal metabolomics, in a known condition, like systolic heart failure or heart failure with reduced ejection fraction?

Dr. Kavita Sharma:

Sure. Well thank you, Carolyn, for the opportunity to chat around this topic. And it's great to be with you and Svati this morning. Metabolomics is a broad general study of essentially, all the chemical processes involving metabolites, or small molecule substrates, their intermediates, and even the products of cellular metabolism. This can be studied in really, any organ system, in any organ. What is really unique, I think, to this particular paper in our project is that, it has yet to have been defined or described in human HFpEF from the myocardial tissue. We call this heart failure with preserved ejection fraction, and inherent to that name in this complicated syndrome is that, there is something probably wrong with the heart, yet we have not really had much insight to what that might be from direct myocardial tissue.

We are also still learning about what metabolomics looks like in, for example, the heart failure with reduced ejection fraction state. Though, there is more published in this space than in HFpEF. From the limited knowledge that we have, it does appear that heart failure with reduced ejection fraction hearts, and this is certainly seen in the plasma, which is where most of metabolomic studies have generated from, those hearts tend to utilize various forms of energy banks, if you will. Whether that's fatty acid oxidation, whether that is glucose utilization or intermediates and so on. And our primary interest was to understand, how do the preserved EF parts in patients fare in comparison?

Dr. Carolyn Lam:

Oh, thank you so much, Kavita. That was beautifully explained. And indeed, what's so special about your paper is, it's not just circulating metabolites but myocardial metabolites. And you have the control groups that are so important to study at the same time. So patients with HFpEF, but also those with HFrEF and versus controls. And thank you for establishing too, that if I'm not wrong, fatty acid metabolism accounts for the majority of ATP generation in the normal heart. Whereas, this declines a little in the HFrEF heart. And now, I think we're about to find out what happens in the HFpEF heart. So if you could explain what you did and what you find.

Dr. Kavita Sharma:

Yes, absolutely. So we examined, again, tissue and plasma metabolomics from 38 subjects with HFpEF. These are patients referred to the Hopkins HFpEF Clinic. And so they have been essentially, clinically evaluated, and have what we define as HFpEF, based on hemodynamic testing. So a right heart catheterization, often with exercise, that meets criteria for diagnosis of the syndrome.

As you stated, we compared our HFpEF patient tissue and plasma samples to samples coming from patients with HFrEF, dilated cardiomyopathy, and non-failing controls. And the latter two sources were a tissue bank from the University of Pennsylvania, that is long-standing, where patients with endstage dilated cardiomyopathy are able to have tissue banked at the University of Pennsylvania at the time of explant prior to transplant. So albeit, we are comparing to fairly advanced end stage dilated cardiomyopathy, and control tissue comes from unused donor hearts, essentially. So presumably, normal heart function patients, likely in a brain death state, who for whatever reason, the hearts were not utilized for transplantation. Again, not an entirely perfect controlled state, but again, given the nature of the work, the fact that it's myocardial tissue, the closest that we have found we've been able to come to for a control comparison.

We started out performing what we call quantitative targeted metabolomics. We measured organic acids, amino acids, and acylcarnitines in the myocardium. And that was totaling around 72 metabolites. And we did the same in plasma, so close to 69 metabolites. And our metabolomics work was actually completed at the University of Pennsylvania. And so, I wish to credit Dr. Zoltan Arany and Dr. Dan Kelly for their great collaboration in this study.

Dr. Carolyn Lam:

That's wonderful. Kavita, if you could tell us a little bit more about the patients with HFpEF. We understand it was end stage dilated cardiomyopathy, HFrEF, and donor hearts as the controls, but the patients with HFpEF, in relation to obesity, diabetes, and how that may impact the interpretation of the results.

Dr. Kavita Sharma:

Sure. So these are HFpEF patients that are in an ambulatory state outpatient setting. They have many of the comorbidities we know are intrinsic today to HFpEF. Out of our HFpEF population, the majority were women. So 71%, that's 27 out of the 38 we serve. And we're very fortunate to serve a African-American enriched population in Baltimore that's intrinsic to our center. And so, over half of our patients were Black. The remaining Caucasian, one non-Caucasian. Over half had been hospitalized, for example, in the prior one year. So these are certainly symptomatic patients. And all had NYHA II or greater symptoms.

We do have a rather obese cohort at Hopkins. And so, our median BMI, for example, was 39, our mean is very similar. And the majority have, as we see often in HFpEF, the majority with hypertension, over half with diabetes. In fact, it was actually 70% or so. Rather few with coronary disease, and this is a trend we're seeing in general in HFpEF in the present day kind of common phenotypes, and about a third with atrial fibrillation. So really, representative, I think, of this kind of cardiometabolic as we call it, phenotype of HFpEF, that is the predominant phenotype we're seeing, at least in North America.

Dr. Carolyn Lam:

Oh, that's perfect. And then, maybe just a few words about the results before I bring Svati in for her thoughts. Thanks.

Dr. Kavita Sharma:

Sure, absolutely. So we conducted this study in a couple different stages. We first started with performing a principal component analysis and hierarchical clustering analysis, to see whether the myocardial metabolites and the plasma metabolites, respectively, would they distinguish these three patient groups? So HFpEF from HFrEF and controls. And interestingly, in the myocardial tissue, our PCA analysis and our hierarchical clustering analysis show that actually, in fact, as few as 70 metabolites in the myocardium really distinctly differentiate these three subgroups. The top contributors that separated HF from controls, for example, and HFrEF, were mostly related to amino acids, including branched chain amino acids and their catabolites, as well as medium and long chain acylcarnitines, which are byproducts of fatty acid oxidation.

When it came to the plasma metabolome, on the other hand, there was far less distinguishing between the groups, and significant overlap, both in PCA and hierarchal clustering. And really, the take home there is that, the myocardial tissue and the plasma were really quite distinct for the overall metabolite analysis. But then, even as we broke it down by fatty acid oxidation, by glucose metabolism, and even branched chain amino acids, we saw this trend continue, that the plasma was quite distinct from the myocardial tissue.

Now, which of the two is more representative of the disease state? Which is the one that we should be paying more attention to? I think that remains to be fully understood further. And of course, it would be really nice to replicate these findings in another cohort. But that is something that, I think, is a first, that certainly, that we have seen and important for the community.

Dr. Carolyn Lam:

Indeed. Oh, Kavita, we could go on talking forever, but I'd really love Svati's thoughts. Why was this paper so special? What does it tell us clinically with any implications?

Dr. Svati Shah:

Yeah. I just want to commend Dr. Hahn, Dr. Sharma, on this incredible work. If you can just imagine how much painstaking work this took for Dr. Sharma and Dr. Hahn. It's a very careful phenotyping of HFpEF. These are true HFpEF patients. The ability to get tissue, and to pair the tissue to the plasma, so that we can really understand. When we measure things in the circulation, and we think they're telling us about the heart, are they actually telling us about the heart? So I really want to commend this incredible work.

And Carolyn, I love talking about cardiac metabolism, because the heart is an incredible organ, right? The heart is a metabolic omnivore. It'll eat many different kinds of fuels, and a lot of different things determine which fuels it uses. And as you nicely outlined, Carolyn, earlier, in the normal heart, the heart prefers to use fatty acids.

But what we are not completely certain of is, what happens in HFpEF? So in HFrEF, we know that the heart switches to glucose, which is not a great fuel, actually. It's actually, a metabolically inefficient fuel. And so we know in HFrEF, that the heart has this metabolic inflexibility. All of a sudden, it's not an omnivore, and it's kind of stuck with certain fuels, which are not very healthy for it.

But what Dr. Sharma and Dr. Hahn have shown, for the first time really, is what happens in HFpEF? And so, I think it's really cool that, actually, it just highlights how complex HFpEF is as a disease. So they were able to show that in some ways, HFpEF is similar to HFrEF, including that there's impairments in use of these fatty acids, which is what the normal heart does.

But, they also show that HFpEF may be different than HFrEF in many ways, including, because of these branched chain amino acids. And that may be because of some of the clinical differences that we know exist in patients with HFpEF, including the obesity and diabetes, that Dr. Sharma nicely outlined. Although, I want to point out, they were very careful about trying to take these clinical factors into account when they looked at differences in the metabolites.

So really incredible work, highlighting that the HFpEF heart also has this metabolic inflexibility. It also is not a metabolic omnivore like the normal heart is, but highlighting important differences, potentially, between HFpEF and HFrEF.

Dr. Carolyn Lam:

Oh, Svati, thank you for putting that so clearly.

Dr. Kavita Sharma:

No, I think that was a really elegant summary of the findings, Svati. And thank you for your kind words and support in allowing us to share our work through Circulation. I really couldn't say it better, but that's exactly what we seem to find is that, when we look at various sort of stores or banks of energy resource, what we really found is that these HFpEF hearts are energy inflexible, as Svati said, that begins with fatty acid metabolism. And so, when we look at, for example, medium and launching acylcarnitines, what we find is that these are markedly reduced in HFpEF myocardial tissue, quite similar to HFrEF. Again, both of them reduced compared to controls. And again, these are byproducts of fatty acid oxidation, and that is really responsible for almost 80% of generally what we think of energy metabolism in the normal state.

In the plasma, however, again, back to that theme where we don't see that reproduced in the plasma, we find that HFpEF is actually not too dissimilar from controls for certain medium and long chain acylcarnitines, and then closer to HFrEF in some cases. And interestingly, we compared our metabolomics study to our prior report of our RNA sequencing paper, that was also published in Circulation now two years ago. And what we found is that, there is reduced gene expression of many of the proteins involved with fatty acid uptake and oxidation, when we compare them to control states. So the story is sort of, fits with what we have seen previously, and when we focus in on this group of genes.

Our analysis of glucose metabolism though, did not include glycolysis or glucose oxidation intermediates. We still found that, majority of the TCA cycle intermediate, so succinate, for example, fumarate, malate, were all reduced in HFpEF versus control. It was really only pyruvate in isolation that was increased in HFpEF myocardium, compared to controls. And again, a number of genes implicated in glucose metabolism in general, we found to be lower in HFpEF versus control, including gluten 1, or SLC2A1, which is involved in glucose uptake.

So again, this theme of, we have patients with significant obesity, many in the diabetic state, we would think that these hearts would utilize these energy stores, but they don't seem to be. And finally, we see distinct differences in the tissue and branched chain amino acid pathways as well. There appears to be some sort of a block between the branched chain amino acids, and then sort of byproducts, as you continue down through ketoacids and further. And we don't fully understand where those blocks are, but that was certainly notable.

And then lastly, I'll say, one interest that we've had, and really, what led to much of this work in the tissue, is to pursue what we call deep phenotyping. Can these molecular signatures, whether it's gene expression, or metabolomics, or what we're working on now, which is proteomics, can these really help us identify unique subgroups within HFpEF? And so, we've tried to do that with the metabolomics, and we found that, using various sort of clustering analytical methods, in fact, there is significant overlap, as it turns out, within HFpEF, when it comes to the metabolomic signatures. And we only found, really, two subgroups within HFpEF. And even these two really did not have much that distinguished them, beyond branched chain amino acids.

And so, this is the first time, at least that our group has seen, at a tissue level, that there is actually a fair bit of homogeneity now in the metabolomic signatures, compared to our RNA sequencing work. And that may be reflective of now, this increasingly cardiometabolic phenotype of HFpEF. And now, we may be seeing signs of that at the clinical and at the treatment level, where we have therapies like SGLT2 inhibitors, that are showing benefit to what seems to be a much broader spectrum of HFpEF, compared to prior therapies. So a lot of questions that have been generated from the work, and we're looking forward to exploring much of this in more detail.

Dr. Carolyn Lam:

And Svati, may I give you the last word? Where do you think this field is headed next?

Dr. Svati Shah:

I think there's so much to do, and I think Dr. Sharma and Dr. Hahn have highlighted how much work there is to do in this space. We're brushing the surface and understanding cardiac metabolism with this really important paper. But Carolyn, as you pointed out, we really need to understand what happens to these patients over time? What happens to, not just cardiac metabolism, but molecular biology more broadly, in patients with HFpEF with these various treatments? Including now, thank goodness, we have SGLT2 inhibitors as a therapeutic intervention for patients with HFpEF. And in fact, we published in Circulation a few months ago, a paper led by a very talented junior faculty, Senthil Selvaraj, where we actually showed that these acetylcarnitine levels that reflect fatty acid oxidation actually are changed by SGLT2 inhibitors, and are associated with changes in clinical outcomes in HFpEF. So we really need larger sample sizes, being able to look at these patients in a longitudinal fashion. But really, doing what Dr. Sharma and Dr Hahn have done, which is careful, careful phenotyping and multidisciplinary teams, so that we can understand the molecular biology, as well as the clinical implications.

Dr. Carolyn Lam:

Oh, wow. Thank you so much, Kavita and Svati, for this incredible interview. I learned so much, and enjoyed it so thoroughly, as I'm sure our listeners did as well.

Well, listeners, you've been listening to Circulation on the Run. Thank you for joining us today, and don't forget to tune in again next week.

Dr. Greg Hundley:

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, please visit ahajournals.org.