Interview | This year, there have been landmark clinical trials that have left cardiologists with some major therapeutic advances for patients with heart failure and reduced ejection fraction. But there have been less advances for neuro-hormonal modulation device therapy in terms of heart failure with preserved ejection fraction or HFpEF, and specifically the inclusion criteria of trial subjects. CSWN spoke with Jacob Kelly, MD, FACC, a fellow of cardiovascular disease at Duke University Medical Center, and first author of a state-of the-art review, published in a recent issue of JACC, “Patient Selection and Heart Failure with Reserved Ejection Fraction.”

CardioSource WorldNews: It’s an interesting story. Tell us about the paper first.

Jacob Kelly, MD: Thank you so much for this opportunity to represent Duke University and Global Cardiovascular Clinical Trialists—and all my colleagues who helped with this paper. Review papers are kind of an interesting animal in that you’re trying to look at all the available evidence and to talk to your audience. So I think it’s important to understand this process. Specifically, for this paper, it focused on the discussions between an important group of folks from the global cardiovascular clinical trial list that’s been forming for 10 years. That’s led by Faiez Zannad, MD, and Bertram Pitt, MD, and it brings together clinician scientists, clinical trialists, regulators from Europe, all around the world, the United States. I think what they do is to critically appraise very recent—and even compared to older trials—to discuss how we can build future clinical trials that are actually successful at treating our patients.

Why is it important to better understand the methodologies used in these HFpEF trials?

That’s a great question. When you think about how a clinical trial is built you need to identify your patients, you need to identify a therapy that may be benefitting these patients, and you have to figure out how you are going to decide if that worked or not. So by looking at the methodology of selecting patients in previous clinical trials and by looking at their successes and/or failures, you can then decide what’s been helpful, use those lessons, learn from them, and design future trials that can best evaluate if a therapy (such as LCZ of the recent PARADIGM trial that’s been so successful in HFrEF) may be beneficial in HFpEF. But if you do the trial wrong and you get intermediate values, you may not have a conclusive answer. So that’s why the methodology is so important. You have to select your patients appropriately, and that’s what we try to understand in this review and discuss with our readers and our clinicians.

And what you found is that it varies.

Absolutely. I think it’s important to understand your patient population. You look at the patients whom you want to study. You look at the intervention or the therapy—be that a drug or perhaps a device—and then you need to mesh the things together. By looking a little bit at prior trials, looking at those patient-selection criteria that drove event rates, you can get an idea of how to power your study and how to select patients so that you have your best chance of determining if that test, that therapy is helpful for our patients, which is really what we want. We want to improve the lives of our patients.

You had to wrap your mind around the fact that you’ve got guideline definitions of HFpEF, you have clinical trial definitions, and you have cut points used to indicate HFpEF that can vary—you had some real challenges here. What did you find when you put it all together?

You need to take this group of thought leaders in the field and see what they’ve said about previous trials—critically appraise and think about it, and see where things have been successful. What we found, just on our review—and this is what we suggest, and you can see that from the trial evolution from previous to current—that by using higher levels of EF, you’re selecting for a more true HFpEF versus a HFrEF population by using elevated levels of naturetic peptide levels, even higher perhaps in patients with co-morbidities, which are very common and very important in HFpEF, such as atrial fibrillation; you can select for specificity of HFpEF. Then look at how the patient’s prior history of heart failure is defined, and how they are being defined now. Sometimes you need to use things like hemodynamics or stress testing to really confirm that HFpEF is HFpEF. Using this whole milieu of patient selection criteria and looking at where they have driven event rates and prior trials is how you can successfully define a trial that’s going to be successful for your therapy at play.

It’s largely subjective in terms of where we are right now. You’re hoping to make that a little less subjective.

I think that’s the future. We still have to focus on patient-selection criteria from clinical definitions by slowly peppering in and adding in some of the omics data, some of these phenomapping mapping techniques that Sanjeev Shah, MD, has done, studied, demonstrated, and published in JACC on how to use novel diagnostic techniques to identify these patients. By utilizing the actual definitions that clinicians are using to define their patients in clinical trials, we’re going to inform clinicians of how to treat these patients. If we don’t use what the actual clinicians are using as definitions, then it’s not going to inform or change treatment of patients. This is how we’re defining HFpEF out there. It’s very subjective, so we try to solidify it the best we can so that we can give that message to clinicians, but also study it in a rigorous way so that we know if when we find that therapy that works for these patients—and we will—that we can move that message to clinicians and into practice.

With the methodologies being all over the map, it’s no wonder that it’s hard to find treatments that work. We really need a better understanding of who these people are and then study them very carefully.

That’s exactly right. By using current clinical standard definitions that are universal from country to country, site to site, from language to language, and then using and adding on some of these new omics techniques or imaging techniques, we can better understand the population and better choose the right patient population to treat.

I also think it’s very important to identify the folks who have been so instrumental at CVCT, as well as in mentoring fellows so that we can be the future clinical trialists. Without Robert Mentz, MD, FACC, and Carolyn Lam, MBBS, FACC, along with the leadership of Dr. Zannad and Pitt, the future of clinical trials would not be so strong.

Kelly, JP, Mentz RJ, Mebazza A, et al. J Am Coll Cardiol. 2015;65(16):1668-82. doi:10.1016/j.jacc.2015.03.043.

Keywords: Population, Clinical Trials as Topic, Patient Selection, Cardiovascular Diseases, Heart Failure, Stroke Volume

< Back to Listings