Author + information
- Joel D. Schilling, MD, PhD∗ ()
- Diabetes Research Center, Department of Medicine, Department of Pathology and Immunology, Washington University School of Medicine, St. Louis, Missouri
- ↵∗Address for correspondence:
Dr. Joel D. Schilling, Division of Cardiology, Diabetes Research Center, Washington University School of Medicine, 660 South Euclid Avenue, St. Louis, Missouri 63110.
Heart failure with preserved ejection fraction (HFpEF) remains a challenging clinical syndrome to diagnose and treat. To date, the majority of clinical trials using therapeutics that are effective for heart failure with reduce ejection fraction (HFrEF) have not produced convincing benefits in those with HFpEF. Although there are many potential reasons for these disappointing results, one clear contributor is the pathophysiologic heterogeneity that exists within the “HFpEF basket.” Several risk factors have been associated with HFpEF, including age, sex, hypertension, metabolic disease, renal dysfunction, and pulmonary hypertension; however, it has been challenging to disentangle the role of these co-morbid conditions at the level of individual patients.
A significant advance in our understanding of HFpEF heterogeneity came from an unsupervised clustering analysis of clinical/laboratory, imaging, and hemodynamic data from patients with HFpEF. This study identified 3 distinct phenotypes of HFpEF in an unbiased way and highlighted the potential importance of accounting for heterogeneity when exploring therapeutic strategies (1). Using a similar approach, a more recent study analyzed patient phenotypes from clinical data, imaging, and blood samples available from the TOPCAT (Treatment of Preserved Cardiac Function Heart Failure with an Aldosterone Antagonist) trial, a randomized study of spironolactone in HFpEF (2). Using these parameters, the investigators could also discern 3 distinct HFpEF phenotypes. In both of these studies, there was an HFpEF phenogroup that was characterized by obesity, diabetes, and elevated inflammatory markers. Interestingly, patients with this metabolic heart disease (MHD) phenotype had a worse prognosis and better response to treatment with spironolactone compared with the other HFpEF subsets in the TOPCAT dataset. This provocative post hoc analysis provided proof-of-concept evidence that patients with distinct HFpEF phenotypes might respond differently to heart failure therapeutics.
The study by Croteau et al. (3) in this issue of JACC: Basic to Translational Science represents an important advancement in our understanding of disease therapy and mechanisms for the metabolic phenogroup of HFpEF. The authors used a mouse model of MHD to explore the impact of angiotensin receptor blockade (valsartan [VAL]) with or without neprilysin inhibition (sacubitril [SAC]) on myocardial disease phenotypes. A mouse model of MHD was generated by feeding animals a high-fat, high-sucrose diet for 4 months, which produced left ventricular hypertrophy (LVH), diastolic dysfunction, perturbed myocardial energetics, oxidative stress, and mild myocardial fibrosis, all of which are features of the human disease. In addition to diet, mice were treated with vehicle, VAL alone (VAL), or VAL with SAC (SAC/VAL). The primary findings were that both VAL and SAC/VAL improved LVH and myocardial energetics, but only mice treated with SAC/VAL had improvements in diastolic function, oxidative stress, and myocardial fibrosis.
A major strength for the translatability of this study (3) was the use of oral administration and careful dose validation for SAC/VAL in which plasma neprilysin activity was shown to be inhibited by >80%. The findings that LVH and myocardial energetics were improved to a similar extent with both VAL and SAC/VAL but fibrosis and oxidative stress were more dramatically impacted with SAC/VAL suggest that neprilysin inhibition has unique effects on myocardial remodeling in MHD. Interestingly, the fact that diastolic function was only improved by the combination of SAC/VAL argues that oxidative stress and fibrosis may contribute more directly to dysfunctional myocardial relaxation with metabolic syndrome. The antifibrotic potential of combination therapy is particularly intriguing. Similar results were found in a mouse model of diabetes and ischemia reperfusion injury in which combination therapy reduced myocardial fibrosis compared with vehicle control or VAL alone (4). In both studies, the combination of SAC/VAL more potently suppressed collagen production at the transcriptional level and diminished interstitial fibrosis at the myocardial level. Whether increased levels of natriuretic peptides account for this antifibrotic phenotype, as has been shown for brain natriuretic peptide, remains to be proven (5).
Although the study by Croteau et al. (3) suggests that SAC/VAL may have unique benefits in MHD, there are some important limitations that must also be considered. The use of a high-fat, high-sucrose model is a reasonable approximation of the “western” diet as it contains high levels of fat, cholesterol, and carbohydrates; however, it is not a model of heart failure. Rather, it more likely approximates early stages of MHD, and therefore whether these findings will translate to humans with more advanced disease is unclear. Another important consideration is that the mice in this study were started on treatment at the same time the diet was initiated. Therefore, the data presented reflect the use of these medications as a preventative rather than as a treatment strategy. Demonstrating similar findings in mice with established metabolic syndrome would enhance the relevance of the data to patients with established obesity and heart disease. The authors also did not collect or present data about the effects of VAL and SAC/VAL on metabolic parameters such as weight gain, insulin resistance, hepatic steatosis, or inflammation. Without these data, it is hard to determine whether the beneficial results with SAC/VAL are directly related to its effects on the myocardium and vasculature or secondary to an improvement in systemic metabolism and/or a reduction in inflammatory cytokine release.
The larger question is how we should interpret the findings of this study (3) in light of the recently published PARAGON-HF (Efficacy and Safety of LCZ696 Compared to Valsartan, on Morbidity and Mortality in Heart Failure Patients With Preserved Ejection Fraction) trial that compared VAL versus VAL/ SAC in patients with HFpEF (6). Although the primary outcome of the PARAGON-HF trial was not met, there was a trend toward reduced heart failure hospitalizations with combination therapy. In most circumstances, a negative result in a definitive clinical trial would suggest that further basic science research into the area is not warranted. In light of this, why should we care about new data investigating SAC/VAL in a mouse model of heart disease? To answer this question, it is important to first consider the details of PARAGON-HF. In this trial, efforts were made to enroll a pure HFpEF population, which was defined as patients with a left ventricular ejection fraction >45%, elevated natriuretic peptide levels, and structural heart disease. However, as with all previous HFpEF trials, there were no attempts to focus on specific HFpEF phenotypes, likely due to concerns about enrollment. The population included in this trial was 50% female, 44% diabetic, and had an average age of 73 years. In subgroup analysis, there was a suggestion that those with lower LVEF and women may derive more benefit from combination therapy. Of note, patients with a history of diabetes did not have a differential response to treatment with VAL/SAC.
There are several important implications of the current study (3) that should be considered in the context of the results from PARAGON-HF (6) and our understanding of HFpEF heterogeneity. First, the pre-clinical study focused entirely on the cardiac effects of SAC/VAL in MHD. The findings supported the possibility that the metabolic phenogroup of HFpEF may respond better to this treatment approach, analogous to that seen in the post hoc analysis of TOPCAT (2). Moreover, with the mouse model, there was no evidence of heart failure, and therefore it is actually a model of obesity-induced cardiac remodeling rather than HFpEF per se. Thus, the favorable effects of combination therapy may be most impactful when initiated early in disease. Third, the aging process can also influence both metabolic disease and cardiovascular pathology. The patient cohort in PARAGON-HF largely consisted of elderly patients who likely had more advanced disease. In contrast, the pre-clinical model used mice of early to middle age. Performing additional translational experiments on animals of advanced age fed a high-fat, high-sucrose diet would shed light on the impact of this variable on disease progression and the response to SAC/VAL. Lastly, the issue of sex needs to be considered. In PARAGON-HF, female subjects seemed to derive more benefit from combination treatment. However, most animal models of metabolic disease use male mice because they develop more profound metabolic syndrome with less time on a high-fat diet. Therefore, to augment the translatability of pre-clinical models, it will be necessary to incorporate sex and age into experimental design.
The syndrome of HFpEF has perplexed many cardiologists and cardiovascular researchers over the past 2 decades. More recently, significant strides have been made in understanding the heterogeneity of this condition. Although the data from PARAGON-HF (6) insinuate that we should give up the ghost of applying HFrEF therapies to HFpEF, the study from Croteau et al. (3) provides a strong counterpoint against this notion. Specifically, the authors provide evidence that MHD may respond favorably to SAC/VAL and that the addition of SAC seems to specifically drive antioxidant and antifibrotic pathways in the heart. Given that inflammation is tightly linked to oxidative stress and fibrosis and is a defining feature of the obesity phenogroup of HFpEF, it is attractive to speculate that modulation of inflammatory pathways may be part of the mechanism for this observation. Ultimately, a key concept that is further emphasized by this article is that phenotypic subsets of HFpEF may respond differently to therapeutics. Despite the challenges this poses for clinical investigation, future studies will need to account for heterogeneity within the syndrome of HFpEF if effective treatments are to be discovered.
↵∗ Editorials published in JACC: Basic to Translational Science reflect the views of the authors and do not necessarily represent the views of JACC: Basic to Translational Science or the American College of Cardiology.
Dr. Schilling is supported by the National Institutes of Health (grant R01 DK11003401) and the American Diabetes Association (grant 118-IBS280).
The author attests he is in compliance with human studies committees and animal welfare regulations of the author’s institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information, visit the JACC: Basic to Translational Science author instructions page.
- 2020 The Author
- Shah S.J.,
- Katz D.H.,
- Selvaraj S.,
- et al.
- Cohen J.B.,
- Schrauben S.J.,
- Zhao L.,
- et al.
- Croteau D.,
- Qin F.,
- Chambers J.M.,
- et al.
- Tamura N.,
- Ogawa Y.,
- Chusho H.,
- et al.
- Solomon S.D.,
- McMurray J.J.V.,
- Anand I.S.,
- et al.