Vive les Differences!-A case for optimism in the treatment of patients with heart failure and preserved ejection fraction?

Abstract

Heart failure with preserved ejection fraction: a poorly understood heterogenous condition The toast ‘Vive la difference!—Long live the difference!’ is routinely used to celebrate individual variation, particularly of men and women.1 However, arguably the phrase promotes conformity to male and female stereotypes, ignoring the wide spectrum of abilities and diverse characteristics within each gender group. There may be equally large differences within each gender than between the sexes? How does this relate to patients with heart failure with preserved ejection fraction (HFPEF)? It's over a quarter of a century since the original description of this condition as a syndrome distinct from the more traditional heart failure with reduced ejection fraction (HFREF).2 The (singular) original difference between both syndromes was that the left ventricular ejection fraction was normal in HFPEF but reduced in HFREF. The causes, precise pathophysiology and diagnostic criteria of HFPEF have been contentious ever since and remain so to this day. It does seem clear, however, that there is a distinct category of the clinical heart failure syndrome which differs from HFREF in many ways though the prognosis is similarly poor. The most painfully obvious difference is that while many treatments have been found in randomised controlled trials (RCTs) to improve outcomes and prolong life in patients with HFREF, nothing has so far been shown to enhance survival in patients with HFPEF.3, 4 Large RCTs have consistently shown that angiotensin converting enzyme inhibitors (ACEi) and angiotensin receptor blockers (ARB) improve survival in patients with HFREF.5-8 This is largely consistent with data from most large observational studies.9-12 In contrast, two large RCTs of ARB13, 14 and one moderately large study of ACEi15 showed no survival benefit in patients with HFPEF. As a consequence of these and other trials, HFPEF is often regarded as a disease in which symptom control is dependent upon the use of diuretics but in which prognosis cannot be modified; a near untreatable condition.4, 16 However, several observational studies have reported a significant survival benefit of ACEi and ARB use in patients with HFPEF.12 In particular, two large studies, using propensity score matching to minimise the risk of confounding,17, 18 have reported significant benefit from ACEi and/or ARB in patients with HFPEF.19, 20 In this issue of IJCP, Egido and colleagues provide a third large study of 3864 patients with HFPEF in which the use of ACEi and/or ARB was associated with reduced all-cause and cardiovascular mortality and lower hospitalisation rates after adjustment for the propensity to take ACEi/ARB and other medications, comorbidities and other potential confounders.21 Why the difference? The biases and limitations of observational studies have been long-recognised with residual confounding from known confounders after adjustment and confounding from unknown or unrecorded variables.12, 22, 23 Thus, RCTs, or meta-analyses of RCTs remain the highest form of evidence and significant change in practice should, wherever possible, be based on evidence from rigorously conducted RCTs.12, 23 However, RCTs in HFPEF so far may not have been as overwhelmingly ‘negative’ or as conclusive as commonly believed. The Perindopril in Elderly People with Chronic Heart Failure (PEP-CHF) RCT compared perindopril 4 mg with placebo in 850 patients aged over 70 with heart failure diagnosed in the previous 6 months with an LVEF >40% and at least three echocardiographic markers of diastolic dysfunction.15 Both recruitment and event rates were lower than predicted and the study terminated once all patients had reached 1 year of follow-up. The primary end-point of all-cause mortality was negative (Hazard ratio [HR] 0.92 95% CI 0.70-1.21) but the study only had 35% statistical power to detect a difference in mortality. By 1 year, 28% of patients in the perindopril arm and 26% in the placebo arm were taking open-label ACEi further diminishing power. Despite this, there were significant reductions in heart failure hospitalisations and improvements in symptoms and exercise capacity in the perindopril-treated group. Similar findings have been reported from the two large RCTs with ARBs. In the Effects of Candesartan in Patients with Chronic Heart Failure and Preserved Left Ventricular Ejection Fraction (CHARM-Preserved) trial,14 the primary end-point of cardiovascular death or hospitalisation for heart failure was not significantly different (adjusted HR 0.86; 95% CI 0.74-1.00), secondary end-points of at least 1, or recurrent, heart failure hospitalisation were positive.24 In the Irbesartan in Patients with Heart Failure and Preserved Ejection Fraction (I-PRESERVE) trial, although the primary end-point of all-cause mortality or hospitalisation for a cardiovascular cause was not significantly different (HR 0.95 [95% CI 0.85-1.05]), 39% and 40% of patients in the placebo and treatment arms took an ACEi at some point during the trial. The Treatment of Preserved Cardiac Function Heart Failure with an Aldosterone Antagonist trial (TOPCAT)25 trial randomly allocated 3445 patients with heart failure and an LVEF >45% from six countries to spironolactone or placebo. Although the primary outcome, a composite of cardiovascular death, aborted cardiac arrest or heart failure hospitalisation was not different, there was a significant reduction in heart failure hospitalisation in the treatment group.25 Furthermore, subsequent analyses revealed that patients from Russia and the Republic of Georgia had much lower event rates,26 many probably did not have HFPEF26 and were much less likely to have taken the study drug.27 Reanalysis excluding patients from these two countries showed significant decreases in the primary and secondary end-points with spironolactone treatment.26 Other potential explanations for the disconnect between RCTs and observational data in HFPEF revolve around the use of ‘real world’ data. Patients with HFPEF are generally elderly people with multiple comorbidities.28, 29 Many of these comorbidities exclude patients from RCTs and therefore the patients with HFPEF who participated in the clinical trials may not be genuinely representative of patients treated in routine clinical practice.30 The wide phenotypic and pathophysiological heterogeneity of HFPEF may also explain the apparent disconnect between the results of RCTs and observational studies.31-35 Even before the first descriptions of HFPEF as a distinct condition, several subgroups had already been reported in relation to coronary heart disease,36 hypertrophic cardiomyopathy,37 idiopathic restrictive cardiomyopathy,38 valve replacement39 and uraemia.40, 41 Argument has continued about the presence of more sensitive markers of systolic dysfunction than ejection fraction in HFPEF and about the need for markers of diastolic function to make the diagnosis. While the syndrome is readily recognised by clinicians, it has been frustratingly difficult to agree uniform diagnostic criteria for entry in to trials. The use of N-terminal pro-B type natriuretic peptide (NT-proBNP) and of objective echocardiographic markers of structural and functional abnormalities has been a major step forwards but came too late for use in the large trials.42 Indeed, ongoing studies such as the Prospective Comparison of Angiotensin Receptor Neprilysin Inhibitor Global Outcomes in Heart Failure with Preserved Ejection Fraction (PARAGON-HF) trial and the Spironolactone Initiation Registry Randomized Interventional Trial in Heart Failure with Preserved Ejection Fraction (SPIRRIT-HF) trial have set more objective definitions of HFPEF including NT-proBNP levels.43, 44 Therefore, a greater understanding of this phenotypic heterogenicity, including aetiologic and pathophysiologic heterogenicity of the syndrome, could potentially lead to more targeted and more successful clinical trials. Our understanding of the diverse disease pathogenesis including obesity, hypertension, diabetes and chronic kidney disease and the pathophysiological mechanisms that lead to myocardial dysfunction perhaps by driving endothelial dysfunction, coronary microvascular dysfunction and diffuse fibrosis via inflammation and endothelial dysfunction will no doubt continue to evolve. This should lead to targeted therapies to be tested in RCTs. Such a ‘Precision Medicine’ based approach has led to significant improvements in cancer care for example.45, 46 The clinical syndrome of HFPEF should not be seen as an untreatable condition. Several existing treatments may be of benefit to specific subgroups of patients. The future management of these patients, as well as the design of clinical trials, will require nuanced, phenotypic/aetologic/pathophysiologic approaches instead of the historic one-size-fits-all cardiological approach.16 Perhaps, we should no longer celebrate ‘Vive la difference!’ when it comes to HFREF and HFPEF. They may be part of a continuous spectrum of the clinical syndrome of heart failure. There may be greater diversity within patients diagnosed with HFPEF than between HFPEF and HFREF. ‘VIVES LES DIFFERENCES! –LONG LIVE THE DIFFERENCES!’. None.