Original Article Heart Metab.

(2016) 71:4-8

Heart failure with preserved ejection

fraction (HFpEF): a basic and clinical
perspective
Ajay M. Shah, MD, FMedSci1; Philip C. Chowienczyk, FRCP2
Departments of 1Cardiology and 2Clinical Pharmacology and Therapeutics, Cardiovascular Division,
King’s College London British Heart Foundation Centre of Excellence, London, UK

Correspondence: Professor Shah, Cardiovascular Division, James Black Centre, 125 Coldharbour Lane,
London SE5 9NU, UK
E-mail: ajay.shah@kcl.ac.uk

Abstract
Up to 50% of patients with heart failure have a relatively normal ejection fraction; this is called “heart
failure with preserved ejection fraction” or HFpEF. These patients are more likely to be older, female,
hypertensive, obese, or to have metabolic syndrome or atrial fibrillation. They suffer substantial mortality
and morbidity, but there are currently no treatments available proven to improve outcomes. Abnormal
left ventricular diastolic function is thought to be a major feature of HFpEF, and its detection (or the
detection of cardiac structural abnormalities that predispose to diastolic dysfunction) is a requirement
for positive diagnosis in most diagnostic guidelines. However, noninvasive detection of left ventricular
diastolic dysfunction and its distinction from normal cardiovascular aging may be difficult. Clinical
assessment of patients during exercise or similar stress can be very valuable. The clinical pathophysiol-
ogy of HFpEF is heterogeneous and involves not only diastolic dysfunction but also abnormalities in
heart rate, heart rhythm, microvascular resistance, aortic stiffness, and ventricular-vascular coupling,
resulting in impaired systolic and diastolic reserve capacity upon exercise. The mechanisms underlying
these abnormalities are poorly understood. The phenotype bears similarity to normal cardiovascular
aging and could involve similar abnormalities in inflammation and oxidative stress levels. Abnormalities
in nitric oxide/cyclic guanosine monophosphate (cGMP) signaling also possibly contribute. To develop
effective therapies for HFpEF, more rigorous clinical phenotyping and classification are probably needed
to facilitate a better understanding of the disease pathophysiology and the development of more per-
sonalized treatments. L Heart Metab. 2016;71:4-8

Keywords: diastolic dysfunction; HFpEF; ventricular-vascular coupling

Introduction the next 15 years, particularly in the >65 age group.

It causes substantial mortality and morbidity and

C
hronic heart failure (CHF) affects nearly 6 mil- represents a major disease and socioeconomic bur-
lion people in the United States and similar den. CHF is a systemic syndrome involving the heart,
proportions in other industrialized countries.1 vasculature, kidneys, and other organs, but it de-
The prevalence is projected to rise substantially over velops primarily as a result of diverse acquired and/

4
Heart Metab. (2016) 71:4-8 Shah and Chowienczyk
HFpEF—basic and clinical perspective

assessment of LV diastolic function is not always

Abbreviations easy. Assessment during exercise—when diastolic
cGMP: cyclic guanosine monophosphate; CHF: chronic abnormalities are more evident—is particularly valu-
heart failure; EF: ejection fraction; HFpEF: heart failure able though not routinely performed. The most recent
with preserved ejection fraction; HFrEF: heart failure European Society of Cardiology (ESC) guidelines
with reduced ejection fraction; HYVET: HYpertension have added the presence of elevated natriuretic pep-
in the Very Elderly Trial; LV: left ventricular; NO: nitric tide (NP) levels as an essential requirement for the
oxide; NOS: nitric oxide synthase diagnosis of HFpEF.4 However, not all patients with
HFpEF (including those characterized by definitive in-
or genetic structural and functional abnormalities vasive assessment) have elevated NP levels5; the use
of the heart. Forty percent to 50% of CHF patients of these new criteria may therefore exclude important
have a form of heart failure in which left ventricular subsets of patients with HFpEF.
(LV) systolic function, as assessed by ejection frac-
tion (EF) at rest, is relatively well preserved. This type Risk factors and heterogeneity
of heart failure has come to be termed heart failure
with preserved EF (HFpEF). The outcome of patients Whether HFpEF and HFrEF are distinct conditions or
with HFpEF is on average slightly better than for part of a spectrum has been debated, but it is clear
those with reduced EF (HFrEF), but they still have that patient characteristics differ between the groups.
substantial morbidity and mortality, eg, 23% mortal- HFpEF is especially common in the elderly. Patients
ity over 3 years in a large meta-analysis.2 Further- with HFpEF are more likely to be female, have hyper-
more, the prevalence of HFpEF relative to HFrEF is tension, obesity, metabolic syndrome, diabetes, atrial
rising. Although many trials have had limited power, fibrillation, and to lead a sedentary lifestyle than those
treatments used for patients with HFrEF (eg, inhibi- with HFrEF, and they are less likely to have ischemic
tors of the renin-angiotensin-aldosterone system, heart disease. Transition of HFpEF to HFrEF may
β-adrenergic blockers, biventricular pacemakers, largely occur only in those who develop myocardial
and implantable defibrillators) have not been shown infarction.
to reduce mortality in HFpEF. Current treatment is There is significant phenotypic and probably
thus focused on comorbidities.3,4 To develop effec- pathophysiologic heterogeneity among patients with
tive specific treatments, there is a compelling need HFpEF. An important question is whether defining
to better understand the pathophysiology of this more homogeneous subpopulations might allow a
type of heart failure. better understanding of the underlying pathophysiol-
ogy and identification of groups that respond favor-
Diagnosis of HFpEF ably to specific therapies. This idea is supported by
a recent unbiased clustering analysis of nearly 400
The diagnosis of HFpEF is more difficult than HFrEF carefully diagnosed HFpEF patients; in the analy-
and more likely to be inaccurate. It is generally ac- sis, three distinct patient groups could be identified,
cepted that abnormal LV diastolic function (with or which differed in clinical and cardiac structural/func-
without other cardiovascular pathology, as discussed tional characteristics as well as outcomes.6
later) is a fundamental component of HFpEF. Current
American Heart Association (AHA) guidelines require Clinical pathophysiology
the presence of signs or symptoms of heart failure, a
preserved EF (EF≥50%), and objective evidence of LV HFpEF was historically considered primarily to be a
diastolic dysfunction to diagnose HFpEF.3 Because disorder of LV diastolic function (so-called “diastolic
symptoms may be nonspecific and difficult to distin- heart failure”). Although this is undoubtedly a major
guish from those related to aging and comorbidities, feature, it is now evident that HFpEF results from a
such as obesity, and because EF is preserved, the complex and variable interplay of multiple defects in
demonstration of LV diastolic dysfunction becomes LV hemodynamic and reserve function, including ab-
critical, especially in those without definitive signs normalities of heart rate and rhythm, vascular stiffness
(such as pulmonary edema). However, noninvasive and resistance, and ventricular-vascular coupling.

5
Shah and Chowienczyk Heart Metab. (2016) 71:4-8
HFpEF—basic and clinical perspective

LV diastolic dysfunction in HFpEF usually com- substantial proportion of the heart failure would have
prises both an impairment of active LV relaxation and been HFpEF. Patients with HFpEF also have reduced
an increase in passive (late diastolic) stiffness that to- vasodilatation during exercise, which impairs their abili-
gether increase LV filling pressures. The latter may be ty to increase stroke volume.12 A defining characteristic
especially evident during exercise, highlighting that of HFpEF is therefore vascular dysfunction and abnor-
the problem is essentially one of reserve capacity mal ventricular-vascular coupling, which compromises
(as in CHF more generally).7 The left ventricle is typi- optimal hemodynamic function.
cally concentrically remodeled and hypertrophied but Although EF at rest is normal or only mildly re-
rarely dilated. Concentric LV remodeling and intersti- duced in HFpEF, the preceding discussion indicates
tial fibrosis contribute to elevated diastolic stiffness in that there is significant systolic impairment during ex-
HFpEF, but an important additional factor may be an ercise. Indeed, more sensitive indices of contractile
increased cardiomyocyte sarcomeric stiffness, influ- function than EF readily identify subtle abnormalities
enced by the giant elastic protein titin and its phos- of systolic function even at rest. However, there is a
phorylation status.8 In diabetic patients, increased clear difference between HFpEF and HFrEF, with the
stiffness related to advanced glycation end products left ventricle typically being concentrically remodeled
may be important. Impairment of the myocardial en- and nondilated in HFpEF, whereas HFrEF is charac-
ergetic state on exercise, knowing that energy me- terized by substantial LV thinning, dilatation, and sys-
tabolism is crucial for active myocardial relaxation, tolic impairment at rest.
may also contribute to diastolic dysfunction.9
Chronic elevation in LV filling pressure leads to Disease mechanisms
pulmonary hypertension in many HFpEF patients,
especially upon exercise. An associated impairment The underlying mechanistic basis for the multisystem
of right ventricular function predicts worse outcomes. abnormalities in patients with HFpEF is unknown and
Elevated filling pressures also predispose to atrial fi- may vary among patients. Hypertension is an impor-
brillation, which is poorly tolerated because HFpEF tant factor in many patients. The in vivo dissection of
patients are highly dependent on normal left atrial potential disease mechanisms is quite challenging; at
function to adequately fill the left ventricle. Many pa- the same time, there are few reliable animal models
tients with HFpEF have an inadequate increase in of HFpEF.
heart rate upon exercise, related to abnormalities in Many pathophysiologic abnormalities found in
autonomic balance, which impairs cardiac output typical HFpEF resemble those observed in normal
and exercise capacity. aging (which is a dominant risk factor for HFpEF), al-
Vascular dysfunction is an important feature of HF- beit much more exaggerated. HFpEF could therefore
pEF. An increase in aortic stiffness and central aortic be considered a form of premature cardiovascular
pressure increases vascular hydraulic load on the left aging.13 The mechanisms underlying cardiovascular
ventricle, adversely affects LV relaxation, and contrib- aging are complex and incompletely understood.13,14
utes to chronically increased wall stress that may drive However, there is a consensus that there is a general
concentric LV remodeling.10 Myocardial perfusion re- disorder of multiple aspects of the signaling cascades
serve is often reduced, independent of coronary artery and molecular pathways required to maintain cellu-
disease, which may in part be related to the increased lar and organ homeostasis, driven for example by
vascular load. The potential importance of hyperten- increasingly dysfunctional cellular defense pathways
sion in HFpEF is illustrated by findings from the HYVET and by increased oxidative stress. It has been sug-
(HYpertension in the Very Elderly Trial) study, where in- gested that the cardiovascular aging phenotype is the
dapamide (either with or without perindopril) compared consequence of inflammatory defenses generated by
with placebo reduced fatal or nonfatal heart failure cells in an attempt to limit this molecular and signaling
events by 64% in patients with preexisting hyperten- disorder.14 From this perspective, a proinflammatory
sion who were aged 80 years or more (but did not sig- aging-related milieu may be an important aspect of
nificantly reduce cardiac deaths).11 Although the type the underlying mechanism in HFpEF.
of heart failure was not specifically defined in this study, Recently, it has been proposed that dysregulated
the characteristics of the patients were such that a nitric oxide (NO) signaling is an important contributor

6
Heart Metab. (2016) 71:4-8 Shah and Chowienczyk
HFpEF—basic and clinical perspective

to HFpEF pathophysiology.15 NO generated by NO Summary and future perspective

synthases (NOSs) regulates multiple aspects of car-
diovascular function, including vascular tone, myo- The effective treatment of HFpEF represents a major
cardial function, growth, energetics, and autonomic unmet clinical need. Accurate diagnosis of the condi-
function. Endothelial NOS (eNOS) is expressed in tion is difficult, and the mechanisms underlying the
endothelial cells and to a lesser extent in cardiomy- pathophysiology are poorly understood. HFpEF ap-
ocytes, whereas neuronal NOS (nNOS) is found in pears to be a highly heterogeneous condition in which
nerves, cardiomyocytes, and potentially in vascular different subgroups may require different therapeutic
smooth muscle. NO signaling involves the elevation strategies. Although large randomized clinical trials
of cyclic guanosine monophosphate (cGMP) and have validated many new therapies for cardiovascular
cGMP-dependent protein kinase (PKG) activity or disease and prevention over the last 30 years, includ-
the S-nitrosylation of specific protein targets. The ing new CHF treatments, they have relied on a sim-
potential of PKG to modulate cardiomyocyte stiff- plicity of diagnosis and inclusion criteria that appar-
ness via titin phosphorylation may be particularly rel- ently has not worked well for the HFpEF population.
evant to HFpEF.8 Interestingly, recent studies in small We suggest that a more rigorous and detailed clinical
groups of highly selected HFpEF patients identified phenotyping, diagnosis, and classification of HFpEF
increased cardiomyocyte stiffness in myocardial into more homogeneous subtypes is required in order
biopsies, which was related to a reduced PKG- to advance basic understanding of the pathophysiol-
mediated phosphorylation of titin.16 These authors ogy and develop effective personalized therapies. L
proposed that HFpEF involves systemic endothelial
dysfunction (most risk factors predisposing to HF- REFERENCES
pEF impair endothelial function) involving impaired
1. Mozaffarian D, Benjamin EJ, Go AS, et al. Heart disease and
bioactivity of endothelium-derived NO, for example,
stroke statistics-2016 update: a report from the American
due to increased oxidative stress. It should be noted Heart Association. Circulation. 2016;133(4):e38-e360.
that randomized clinical trials of agents that enhance 2. Meta-analysis Global Group in Chronic Heart Failure (MAG-
GIC). The survival of patients with heart failure with preserved
NO release (eg, the β-blocker nebivolol) or prevent or reduced left ventricular ejection fraction: an individual patient
cGMP breakdown (eg, the phosphodiesterase type data meta-analysis. Eur Heart J. 2012;33(14):1750-1757.
3. Yancy CW, Jessup M, Bozkurt B, et al. 2013 ACCF/AHA
5 inhibitor sildenafil) were unsuccessful in HFpEF.17,18
guideline for the management of heart failure: a report of the
However, ongoing trials are assessing agents that American College of Cardiology Foundation/American Heart
directly stimulate cGMP formation (eg, vericiguat; Association Task Force on practice guidelines. Circulation.
2013;128(16):e240-e327.
NCT01951638). 4. Ponikowski P, Voors AA, Anker SD, et al; The Task Force for
Table I shows the potential pathophysiologic ab- the Diagnosis and Treatment of Acute and Chronic Heart Fail-
ure of the European Society of Cardiology (ESC); developed
normalities that may be influenced by antihyperten-
with the special contribution of the Heart Failure Association
sive agents or by agents that target the NO-cGMP (HFA) of the ESC. 2016 ESC Guidelines for the diagnosis
pathway. and treatment of acute and chronic heart failure. Eur Heart J.
2016;37(27):2129-2200.

Mechanism Effects of antihypertensive treatment Effects of NO-cGMP pathway modulation

Increased filling pressures Reduced by diuretics Some diuretic effect
Increased passive diastolic stiffness Reduced by regression of LVH Reduced titin phosphorylation
(structural) Possible specific effect of spironolactone
Impaired LV relaxation Improved by reduction in afterload Improved relaxation
Systolic dysfunction Improved by reduction in afterload Possibly improved by reduction in afterload
Increased aortic stiffness Reduced Minor effects
Possible specific effect of spironolactone
Increased wave reflection from Reduced reflection (minor) Reduced reflection
conduit arteries
Microvascular dysfunction Reduced peripheral resistance Endothelium-derived NO beneficial
Skeletal muscle dysfunction Possible effect Improved mitochondrial respiration

Table I Pathophysiologic mechanisms in heart failure with preserved ejection fraction (HFpEF) and potential benefits of antihypertensive
treatments and manipulation of the nitric oxide–cGMP pathway.
Abbreviations: cGMP, cyclic guanosine monophosphate; LV, left ventricular; LVH, left ventricular hypertrophy; NO, nitric oxide.

7
Shah and Chowienczyk Heart Metab. (2016) 71:4-8
HFpEF—basic and clinical perspective

5.  Anjan VY, Loftus TM, Burke MA, et al. Prevalence, clinical 12. Borlaug BA, Melenovsky V, Russell SD, et al. Impaired chro-
phenotype, and outcomes associated with normal B-type na- notropic and vasodilator reserves limit exercise capacity in
triuretic peptide levels in heart failure with preserved ejection patients with heart failure and a preserved ejection fraction.
fraction. Am J Cardiol. 2012;110(6):870-876. Circulation. 2006;114(20):2138-2147.
6. Shah SJ, Katz DH, Selvaraj S, et al. Phenomapping for novel 13. Loffredo FS, Nikolova AP, Pancoast JR, Lee RT. Heart failure
classification of heart failure with preserved ejection fraction. with preserved ejection fraction: molecular pathways of the ag-
Circulation. 2015;131(3):269-279. ing myocardium. Circ Res. 2014;115(1):97-107.
7. Borlaug BA, Olson TP, Lam CS, et al. Global cardiovascular 14. Lakatta EG. So! What’s aging? Is cardiovascular aging a dis-
reserve dysfunction in heart failure with preserved ejection frac- ease? J Mol Cell Cardiol. 2015;83:1-13.
tion. J Am Coll Cardiol. 2010;56(11):845-854. 15. Paulus WJ, Tschöpe C. A novel paradigm for heart failure with
8. Linke WA, Hamdani N. Gigantic business: titin properties and preserved ejection fraction: comorbidities drive myocardial dys-
function through thick and thin. Circ Res. 2014;114(6):1052- function and remodeling through coronary microvascular endo-
1068. thelial inflammation. J Am Coll Cardiol. 2013;62(4):263-271.
9. Phan TT, Abozguia K, Nallur Shivu G, et al. Heart failure with 16. van Heerebeek L, Hamdani N, Falcão-Pires I, et al. Low myo-
preserved ejection fraction is characterized by dynamic impair- cardial protein kinase G activity in heart failure with preserved
ment of active relaxation and contraction of the left ventricle on ejection fraction. Circulation. 2012;126(7):830-839.
exercise and associated with myocardial energy deficiency. J 17. Conraads VM, Metra M, Kamp O, et al. Effects of the long-
Am Coll Cardiol. 2009;54(5):402-409. term administration of nebivolol on the clinical symptoms,
10. Kawaguchi M, Hay I, Fetics B, Kass DA. Combined ventricular exercise capacity, and left ventricular function of patients with
systolic and arterial stiffening in patients with heart failure and diastolic dysfunction: results of the ELANDD study. Eur J Heart
preserved ejection fraction: implications for systolic and dia- Fail. 2012;14(2):219-225.
stolic reserve limitations. Circulation. 2003;107(5):714-720. 18. Redfield MM, Chen HH, Borlaug BA, et al. Effect of phospho-
11. Beckett NS, Peters R, Fletcher AE, et al. Treatment of hyper- diesterase-5 inhibition on exercise capacity and clinical status
tension in patients 80 years of age or older. N Engl J Med. in heart failure with preserved ejection fraction: a randomized
2008;358(18):1887-1898. clinical trial. JAMA. 2013;309(12):1268-1277.