Transcript

Dr. Greg Hundley:

Welcome listeners to the month of December, and it is the December 3rd issue of Circulation on the Run, and I'm one of your co-hosts, Dr. Greg Hundley, associate editor, director of the Pauley Heart Center at VCU Health in Richmond, Virginia.

Dr. Shirin Doroudgar:

And I'm your co-host, Shirin Doroudgar, assistant professor at the University of Arizona College of Medicine in Phoenix.

Dr. Greg Hundley:

Wow, Shirin, very interesting feature this week, thinking about some of our cardiac biomarkers, so NT-proBNP and cardiac troponin, as well as cardiac troponin T. Now, which of those three might be associated with seven-year changes in cardiac structure and function in older individuals? And what we're going to do to find out is look into results from the ARIC study. Remember the Atherosclerosis Risk in Communities study. But before we get to answering that question, how about we grab a cup of coffee and jump into some of the other articles in this issue? And how about this week I go first?

Dr. Shirin Doroudgar:

Sounds great. Let's do it.

Dr. Greg Hundley:

So, Shirin, guidelines have suggested screening for atrial fibrillation to enable early treatment and avoid downstream negative clinical events, stroke and things of that nature. However, whether biomarkers could be used to stratify individuals, to risk stratify them is not known. And so this team, in association with corresponding author, Dr. Katrin Kemp Gudmundsdottir from Karolinska Institute, aimed to determine if atrial fibrillation screening potentially enhanced by NT-proBNP would reduce stroke or systemic embolism incidence as compared to a control group and then to determine if it was safe for those with low N-terminal pro-BNP concentrations to forfeit prolonged screening. And so to accomplish this, Shirin, this team randomized in a one-to-one fashion 75 to 76-year-old individuals in Stockholm, Sweden to screening versus control without screening, and NT-proBNP concentrations were measured in a single-lead ECG registered only once if NT-proBNP was less than 125 nanograms per liter, whereas if NT-proBNP was greater than 125 nanograms per liter, participants then underwent prolonged screening and recording of single-lead EKGs four times daily for two weeks. If atrial fibrillation was detected, then treatment was initiated.

Dr. Shirin Doroudgar:

Interesting concept, Greg. Basically testing the efficacy of NT-proBNP to guide continuous screening for atrial fibrillation. And what do they find?

Dr. Greg Hundley:

Right. So Shirin, in total, 28,712 individuals were randomized after exclusion of death and immigration. So 1,395 were in the intervention group and 13,884 in the control group. After a median of 5.1 years, there was no difference in the primary outcome of stroke or systemic embolism between the intervention group and the control group. And then the low NT-proBNP group had significantly fewer strokes and systemic emboli than the control group. And then in the high NT-proBNP group, the risk of stroke or systemic embolism was higher compared to the low NT-proBNP group. So Shirin, in conclusion, in this population-based screening trial for atrial fibrillation using NT-proBNP for screening enhancement, the use of NT-proBNP for screening enhancement was safe in identifying the low-risk participants.

Dr. Shirin Doroudgar:

That's great, Greg. So important findings about biomarkers to stratify individuals.

Now, Greg, our next study is about kidney outcomes. So Greg, chronic kidney disease is common in patients with heart failure. It increases in prevalence as heart failure progresses, and it often complicates treatment, and heart failure patients with chronic kidney disease often have worse outcomes than those without. Now, this coupled with the finding that new pharmacological therapies for heart failure have potentially good effects on kidney function have raised the question of whether formal evaluation of kidney outcomes should become routine in future trials of new treatments for heart failure.

However, Greg, kidney outcomes have been variably defined using non-standardized composite endpoints in key heart failure trials, thus introducing complexity in their interpretation and cross-trial comparisons. So Greg, this team led by the corresponding author, Muthiah Vaduganathan from Brigham and Women's Hospital, examined the effects of steroidal mineralocorticoid receptor antagonists, or MRAs, the angiotensin receptor neprilysin inhibitor, ARNI, sacubitril/valsartan, and sodium-glucose cotransporter-2, SGL2 inhibitors on composite kidney endpoints using uniform definitions in six contemporary heart failure trials.

Dr. Greg Hundley:

Oh, wow, Shirin. So applying standardized kidney endpoint definitions to look at the effects of steroidal MRAs, ARNIs, and SGLT2 inhibitors on renal outcomes in heart failure.

Dr. Shirin Doroudgar:

Right.

Dr. Greg Hundley:

So what did they find?

Dr. Shirin Doroudgar:

So among the 28,690 participants across the six trials with the median age of 69 years old, the proportion experiencing composite kidney endpoint with a more stringent definition of a sustained decline in kidney function, defined by eGFR threshold of 57%, ranged from 0.3% to 3.3%. The proportion of patients experiencing this endpoint with a less stringent definition, so that would be eGFR threshold of 40%, ranged from 1 to 10%. The MRAs, double the risk of composite kidney endpoint, were applying the least stringent definition compared with placebo, and these effects were less apparent and no longer significant with application of more stringent definitions, Greg. ARNI appeared to consistently reduce the occurrence of the composite kidney endpoints irrespective of specific eGFR thresholds applied. The potential benefits of SGL2 inhibitors on the composite kidney endpoint was more apparent when defined by more stringent eGFR thresholds, although none of these effects individually were statistically significant.

So Greg, in conclusion, when applying standardized stringent kidney endpoint definitions, MRAs, ARNI, and SGL2 inhibitors have either neutral or beneficial effects on kidney outcomes and heart failure, and then applying less stringent definitions, increased event rates, but included acute declines in eGFR that might not ultimately reflect the long-term effects on kidney disease progression. So future trials should carefully monitor the definition of the composite kidney outcomes and also should consider the evaluation of heart failure patient populations for kidney risk.

Dr. Greg Hundley:

Very nice, Shirin.

Well, we next transition to another clinical paper, and Shirin, the MINT trial, M-I-N-T, raised concern for harm from a restrictive versus liberal transfusion strategy in patients with acute myocardial infarction and anemia. Now of note, however, type 1 and type 2 MI are distinct pathophysiological entities that may respond differently to blood transfusion. So this analysis, Shirin, performed in collaboration with corresponding author, Dr. Andrew DeFilippis from Vanderbilt University Medical Center, and team sought to determine if the effects of transfusion varied among patients with a type 1 or a type 2 myocardial infarction and anemia, and the team hypothesized that the liberal transfusion strategy would be of greater benefit in type 2 as opposed to type 1 MI. And so to address this question, Shirin, the team compared the rates of death through MI at 30 days in patients with type 1, so 1,460 patients, and type 2, so 1,955 patients, with MI and anemia who were randomly allocated to a restrictive, so that's a threshold of 7 to 8 grams per deciliter or a liberal threshold of 10 grams per deciliter transfusion strategy.

Dr. Shirin Doroudgar:

So Greg, examining the outcomes of those randomized to transfusion thresholds in the setting of type 1 versus type 2 MI, what do they find?

Dr. Greg Hundley:

Right, Shirin. So overall they observed the primary outcome of death through MI in 16% of type 1 MIs and 15% of type 2 MIs. Now, the rate of death through MI was higher in patients with type 1 MI randomized to a restrictive, so it was 18%, versus a liberal, or 13%, transfusion strategy, with no difference observed between the restrictive, so about 15 or 16%, and the liberal, around 15%, transfusion strategies in patients with type 2 MI. And so, Shirin, the concern for harm with a restrictive transfusion strategy in patients with acute MI and anemia raised in the MINT primary outcome manuscript may be more apparent in patients with type 1 MI as opposed to type 2 MI.

Dr. Shirin Doroudgar:

Wow, Greg. Thank you so much for telling us about the MINT trial.

Now, Greg, long QT syndrome is a lethal arrhythmia with nearly 30% of cases caused by rare loss of function variants in the cardiac potassium channel gene KCNH2. Variant classification of this gene is difficult due to lack of functional data. Moreover, variant-based risk stratification is also complicated by heterogeneity in clinical data and incomplete penetrance. So here in this study, the authors wanted to test whether variant-specific information primarily from high-throughput functional assays could improve both classification and cardiac event risk stratification in a large harmonized cohort of KCNH2 missense variant heterozygotes.

Dr. Greg Hundley:

Oh, wow, Shirin. So how did they accomplish this study, and what did they find?

Dr. Shirin Doroudgar:

So Greg, in this study led by co-corresponding authors Brett Kroncke from Vanderbilt University and Jamie Vandenberg from the Victor Chang Cardiac Research Institute in Australia, the authors wanted to quantify cell surface trafficking of 18,796 variants in the KCNH2 gene using a multiplex assay of variant effect or MAVE. They recorded KCNH2 current density for 533 variants by automated patch clamping, and they deeply phenotyped 1,458 patients with KCNH2 missense variants, including cardiac event history and mortality. They correlated variant functional data and long QT penetrance estimates with cohort phenotypes and assessed hazard ratios for cardiac events. Greg, they found variant MAVE, or multiplex assay of variant effect, trafficking scores and automated patch clamping peak tail currents were highly correlated. So the MAVE data was found to provide up to pathogenic very strong evidence classification for severe loss of function variants, and Greg, both functional assays and long QT penetrance estimates were significantly predictive of cardiac events when independently modeled with patient sex and adjusted QT intervals.

However, the MAVE data became nonsignificant when currents and penetrant estimates were also available. And Greg, 20-year event outcomes based on patient-specific sex and adjusted QT intervals were improved with prospectively available penetrant scores conditioned on multiplex assays of variant effect, MAVE, or the attainable APC peak tail current data. So Greg, combining automated patch clamping and quantitative long QT penetrance estimates conditioned on the multiplex assay of variance effect data, that MAVE data, improved prediction of 20-year cardiac event outcomes in this large cohort of KCNH2 variant heterozygotes. So Greg, MAVE scores will facilitate clinical classification of new variants discovered in individuals with suspected long QT as soon as they're observed, and what is also really great about this study is that scores and penetrant estimates are readily searchable at variantbrowser.org for community inquiry and genetic workup of suspected inherited arrhythmias.

Dr. Greg Hundley:

Oh, wow, Shirin. Very interesting that some of these estimates are readily available and readily searchable.

Well, next let's take our listeners to the mailbag, and I'll go first. This week, there are two letters to the editor by Professor Zhang and Sandhu regarding the previously published article entitled “The Clinical Impact of Routine Assessment of Patient-Reported Health Status in Heart Failure Clinic, the PRO-Heart Failure Trial.” And then Joe Kislo has just done a fantastic job, a beautiful Perspective piece entitled “Echocardiography: The AHA in a Hundred Years.” And then there's a nice On My Mind piece from Professor Armbruster entitled “Pitfalls of Choosing a Study Endpoint, Including Cardiovascular Death, in Comparative Clinical Trials.”

Dr. Shirin Doroudgar:

And we have a Research Letter by Professor Christian Sohns [en]titled “Catheter Ablation in Patients with End-Stage Heart Failure and Atrial Fibrillation, a two-year follow-up of the CASTLE-HTx trial.” And then we have a Frontiers piece by Jonathan Lindner [en]titled “In Vivo Cardiovascular Molecular Imaging Contributions to Precision Medicine and Drug Development.” And now let's go on to that feature discussion.

Dr. Greg Hundley:

You bet.

Welcome listeners to our feature discussion today on this December 3rd, and a very interesting study relating to cardiovascular biomarkers. And we have with us today the corresponding author, Dr. Amil Shah from University of Texas Southwestern Medical Center in Dallas, Texas, and also the first author, and you heard from him previously, it's Dr. Peder Myhre from University of Oslo in Norway, and then a favorite, our guest editor today, Dr. Allan Jaffe from the Mayo Clinic in Rochester, Minnesota. Welcome to you all, gentlemen.

Well, Peder, we're going to start today with you. And could you describe for us some of the background information that went into the preparation of this study, of the importance of cardiovascular biomarkers? And then also, what was the hypothesis that you wanted to address?

Dr. Peder Myhre:

Greg, thank you so much, and it is such a pleasure and so much fun for me to be on the other side of the table today and actually be interviewed by you. So this has been a very exciting study that I performed with the senior author, Amil Shah, and it digs into the differences between cardiac troponin T, cardiac troponin I, and NT-proBNP. And what we've known is that these biomarkers are super strong predictors of cardiovascular events. That has been shown in the general population, in patients with risk factors for heart disease and also in patients with established cardiovascular disease. We also know that these biomarkers are cross-sectionally associated with worse cardiac function.

And there are some differences. So we know that the troponins are more reflective of an LV hypertrophy kind of phenotype, with LV mass, et cetera, diastolic dysfunction, while NT-proBNP is more strongly related to wall stress, to LV dilatation, et cetera. So there are some differences there. There are also more and more evidence suggesting that troponin T and troponin I different pathophysiology. Therefore, Greg, we wanted to assess whether these three biomarkers were predictive of changes in cardiac structure and function over time. So we followed individuals for seven years and assessed dissociation between these biomarkers and changes in cardiac structure and function.

Dr. Greg Hundley:

Very nice, Peder. Beautiful description and great lead in for our listeners to turn to Amil Shah.

So Amil, Peder nicely described sort of the background information. How did you include in your study population, and then what was your study design?

Dr. Amil Shah:

Thanks very much, Greg. Yeah, so we studied over 2,000 older individuals who are enrolled in a longitudinal cohort study called ARIC, the Atherosclerosis Risk in Communities study, and we included individuals who were free of cardiovascular disease, so known coronary disease, heart failure, et cetera. These individuals underwent cardiac biomarker measurements and echocardiography at the fifth study visit, at which time the mean age was 74, and they also underwent repeat ultrasound imaging seven years later, at the seventh visit, when the mean age was 81. And so that was our population, and then we used multivariable linear regression models, adjusting for a host of demographic features, clinical risk factors, to look at the association between the cardiac biomarkers at the one-time point and how measures of systolic function, diastolic function, and cardiac structure changed over time.

Dr. Greg Hundley:

Very nice. And so a longitudinal cohort of primarily older individuals in their seventies and following them over time and looking at, I guess, imaging outcome measures. And so, Peder, what results did you observe?

Dr. Peder Myhre:

Yeah, Greg. So we made some interesting findings here, and first of all, we confirmed that all three biomarkers were strong predictors of cardiovascular events, and as we've seen previously, it was primarily a strong predictor of incident heart failure.

Next, we looked at the association with these different echo measures, and what we found was that NT-proBNP was a strong predictor of changes in or worsening of cardiac structure and function, so worse LVEF, worse diastolic function, more LV hypertrophy in those who had a higher NT-proBNP.

Then we saw the same signal for cardiac troponin I. The same signal of higher levels was associated with more progression of their cardiac dysfunction. But you're right, very interestingly, we did not see the same for cardiac troponin T, and that, I think Professor Allan Jaffe will have some comments to later, but this adds to a wealth of literature showing that these biomarkers, troponin T and troponin I, probably reflect different things. I also finally want to highlight that we also found that higher levels of these biomarkers also predicted an increased risk of developing dyspnea. So in fact, Greg, those who had biomarker values above the median were approximately 50% at greater risk of developing significant dyspnea.

Dr. Greg Hundley:

Very nice. And both Peder and Amil, either one, were the results similar for both men and women?

Dr. Peder Myhre:

Yes. So we did not see any difference between sexes, and there were also consistent results across age groups, although these were quite old population with not a very large range of age at the baseline visit.

Dr. Greg Hundley:

And were the thresholds different?

Dr. Peder Myhre:

Yeah, so we did look at different thresholds because we know that men have higher levels of troponin compared to women, and it's actually opposite for NT-proBNP. Women have higher levels than men. So we also looked at six specific thresholds and found the same consistent results.

Dr. Greg Hundley:

Very nice.

Well, listeners, next we're going to turn to our guest editor, Allan Jaffe from the Mayo Clinic in Rochester, Minnesota. And Allan, we appreciate your service as a guest editor at Circulation, and we offer many papers to come across your desk because of your wisdom in this area. Could you perhaps identify for us what attracted you to this paper? And then secondarily, how do we put the results of this study really in the context of the use of biomarkers today for forecasting cardiovascular events in older individuals?

Dr. Allan Jaffe:

Well, first of all, one of the attractions in this particular article is it's extremely well done and they were quite careful. There are problems when one gets to low values, which are what are used in this particular study, of tending to have a lot of overlap between the groups, and this group was at least fairly careful to make sure that they were talking about relatively robust changes. That's important. So the paper was well-done and it highlights some differences between the biomarkers that we'll talk about in just a moment because they may have important pathophysiologic, but there may be some analytic reasons as well that ought to be considered.

This sort of paper in the older individuals begins to fall into what is becoming a very, very important trend in younger individuals and particularly diabetics, which is an understanding that at values well below the 99th percentile, which we all use for myocardial infarction, that there is predictive value of both proponent and the natriuretic peptides. And in point of fact, the diabetes groups now use these very low values to suggest that treatment ought to be initiated because of data such as the CANVAS trial with SGLT2 inhibitors and some of the trials, such as STOP-HF and PONTIAC, which suggested that the therapy and interdigitation with therapy would improve prognosis. So it fits very nicely into that category.

The differences in this particular study is it's an older age group, although I must say, and importantly, it's clear the patients were well-screened because the values for troponin were relatively low. And one of the interesting differences was pointed out by your authors, which is that there's a difference or seems to be a difference between troponin I and troponin T in this regard, and there are two or three different reasons why that could be. It is well known that troponin T tends to predict overall mortality a little bit better than troponin I, and that I tends to be more associated with cardiovascular events per se, which suggests maybe a difference in specificity.

One hint to that is that it's known that troponin T at least can be expressed in patients who have chronic skeletal muscle disease. When we first described that in 2011, people were pretty cynical, but I think it's now well-accepted that in patients who have skeletal muscle disease, that you can see elevations in troponin T, and perhaps since many of those skeletal muscle abnormalities can have a negative prognosis, that may be part of the explanation, and that was suggested by the authors.

There's another one as well, which is if one looks at these two assays carefully, the troponin I assay is somewhat more sensitive, but more importantly has better precision at the low end. So the ability to separate out groups is probably a little bit better with troponin I than with troponin T, and that may be an important difference. Troponin T, the generation that was used, the fifth generation, there is a sixth generation coming that will solve that issue and get more sensitive, and so some of that may wash out as we see these differences in the analytics.

So overall, I think this paper suggests that not only in younger individuals, but in older individuals as well, that there is an opportunity if we use these biomarkers intelligently to identify patients who are at risk, and hopefully the older individuals will be like the younger individuals, which have started to show a signal that we can intervene to the benefit of our patients. Additional trials are necessary both in the younger and in the older, but it starts what is potentially an important initiative to try and prevent disease as opposed to using these biomarkers to define acute disease.

Dr. Greg Hundley:

Very nice, very well stated, and thank you so much, Allan.

Well, listeners, we're going to turn back to our authors, and Amil, we'll start with you. What do you see is the next study to be performed in this sphere of research?

Dr. Amil Shah:

Well, I think maybe not the next, but I think the most compelling next step for this type of research will kind of be exactly what Allan had mentioned, which is to move from these studies that look at prognosis to actually think about approaches to implementation to use these to improve health and prevent disease, and I think that's a large step, but I think that would be the next important step.

Dr. Greg Hundley:

Thank you. And Peder, to you.

Dr. Peder Myhre:

Yeah, thank you so much for those insights, Allan. That was fantastic. And I agree with Amil, and in addition, I think we need to improve the accuracy of these biomarkers, especially if we're going to measure them in younger individuals, as Allan pointed out. So we need more precision in measuring the biomarkers, and we also need studies that it can identify how we should use the biomarkers, in what settings should we measure individuals at risk, or should we measure those with established disease, or should we just use it to risk stratify everyone, even younger individuals. So we need to know more about that, Greg.

Dr. Greg Hundley:

Thank you, Peder.

And then, Allan, maybe just to sort of sum up for us here?

Dr. Allan Jaffe:

Two quick things to say. First of all, there probably is additional information if one were to use serial values, because the change is very likely to augment the predictive accuracy. But I do think we're starting to move to an area where many of us have thought we would go for many, many years but which takes additional data, and that is, at some point in time, we'll all have our baseline troponins and our NT-proBNPs when we have our yearly physicals, and the change over time, and they'll be scrutinized if they're too high. Perhaps we'll do some preventive interventions, and if they change over time, again, we'll consider those preventive interventions. This is another important step toward that eventual goal.

Dr. Greg Hundley:

Super.

Well, listeners, we want to thank Dr. Amil Shah from University of Texas Southwestern Medical Center as the corresponding author, and Dr. Peder Myhre from Akershus University and University of Oslo in Oslo, Norway as the first author, and then also our guest editor today, Dr. Allan Jaffe from the Mayo Clinic in Rochester for bringing us this study highlighting that higher NT-proBNP and cardiac troponin I but not necessarily cardiac troponin T concentrations were associated with greater declines in LV function over seven years in late life, independent of some of the more traditional cardiovascular risk factors.

Well, on behalf of Shirin and also Peder and myself, we want to wish you a great week, and we will catch you next week on the Run.

Dr. Peder Myhre:

This program is copyright of the American Heart Association 2024. The opinions expressed by speakers in these podcasts are their own and not necessarily those of the editors or of the American Heart Association. For more, please visit ahajournals.org.