20 BORLAUG BA Circulation Journal REVIEW Official Journal of the Japanese Circulation Society http://www.j-circ.or.jp Mechanisms of Intolerance in With Preserved Barry A. Borlaug, MD

Approximately half of patients with heart failure (HF) have a preserved ejection fraction (HFpEF), and with the chang- ing age and comorbidity characteristics in the adult population, this number is growing rapidly. The defining symptom of HFpEF is exercise intolerance, but the specific mechanisms causing this common symptom remain debated and inadequately understood. Although diastolic dysfunction was previously considered to be the sole contributor to exercise limitation, recent studies have identified the importance of ventricular systolic, chronotropic, vascular, en- dothelial and peripheral factors that all contribute in a complex and highly integrated fashion to produce the signs and symptoms of HF. This review will explore the mechanisms underlying objective and subjective exercise intoler- ance in patients with HFpEF. (Circ J 2014; 78: 20 – 32)

Key Words: Aging; Diastolic dysfunction; Exercise; Heart failure; Hemodynamics

eart failure (HF) represents an enormous worldwide health problem, afflicting 6 million Americans, 15 Review of Basic Hemodynamics H million Europeans, and quite possibly over 35 mil- Cardiovascular hemodynamics, reviewed in detail elsewhere,25,26 lion Asians.1 Half of patients with HF have preserved ejection are fundamentally dictated by cardiac-specific properties, au- fraction (HFpEF), and with current secular trends in the age- tonomic tone, and external forces that modulate cardiovascular distribution and comorbidity profiles of the adult population, performance (afterload and preload). Afterload can be broadly the prevalence of HFpEF is growing relative to HF with re- defined as the forces that oppose cardiac ejection. Afterload is duced EF (HFrEF) by 1% per year.2 Exercise intolerance is the often incorrectly conceptualized as arterial , but cornerstone clinical expression of HF, and exercise capacity is is more accurately reflected as systolic wall stress or aortic similarly impaired in HFpEF and HFrEF.3 However, in con- input impedance. The use of wall stress is problematic because trast to HFrEF, there is no proven effective treatment for it is potently affected by ventricular properties, as wall stress HFpEF.4 The reasons for this disparity are manifold, but is directly related to chamber dimension and inversely to wall among them, incomplete understanding of the pathophysiol- thickness. The ideal measure of afterload would be com- ogy looms large. Initial mechanistic paradigms focused on left pletely independent of the ventricle. Aortic input impedance ventricular (LV) diastolic dysfunction as the dominant mecha- provides a measure of the total vascular opposition to ejection nism driving HFpEF,5 but numerous recent studies have iden- that is independent of the heart, but is cumbersome to use tified abnormalities in LV systolic function, right ventricular because it is expressed in the frequency domain and is there- (RV) function, chronotropic incompetence, abnormal system- fore difficult to use with time-domain measures of ventricular ic and pulmonary vasodilation, endothelial dysfunction, and function. Effective arterial elastance (Ea) is an alternative abnormalities in the periphery.6–24 To make progress in the time-domain measure that incorporates both mean resistive treatment of HFpEF, these different mechanisms require fur- and oscillatory components of arterial load and is defined by ther study. An important paradigm regarding HFpEF that has the ratio of endsystolic arterial pressure to stroke volume emerged in recent years is that the presence of normal cardio- (SV). Ea is directly related to mean systemic vascular resis- vascular homeostasis at rest in no way guarantees preservation tance (SVR) and (HR) and inversely related to total of reserve capacity under physiologic stressors, the most com- arterial compliance. mon of which in daily life is physical exercise. Accordingly, Preload can be defined by the magnitude of distention or detailed understanding the mechanisms of exercise reserve “stretching” of ventricular myocytes prior to contraction, which intolerance is critical to appreciating the complex pathophysi- dictates the extent of myofiber shortening in the subsequent ology of HFpEF. In this review, I shall review the roles of contraction. Preload is often erroneously conceptualized in cardiac, vascular, and noncardiovascular systems in limiting practice as being equivalent to LV filling pressures (LVFP), exercise capacity in patients with HFpEF. when in fact, the most accurate measure of LV preload is chamber volume prior to the onset of isovolumic contraction

Received August 28, 2013; revised manuscript received October 30, 2013; accepted November 11, 2013; released online December 3, 2013 The Division of Cardiovascular , Department of Medicine, Mayo Clinic, Rochester, MN, USA Mailing address: Barry A. Borlaug, MD, Mayo Clinic College of Medicine, 200 First Street SW, Rochester, MN 55905, USA. E-mail: [email protected] ISSN-1346-9843 doi: 10.1253/circj.CJ-13-1103 All rights are reserved to the Japanese Circulation Society. For permissions, please e-mail: [email protected]

Circulation Journal Vol.78, January 2014 Exercise Intolerance in HFpEF 21

(LV end-diastolic volume, LVEDV). Filling pressures are re- with exertion is, not surprisingly, metabolic demand for O2, lated to preload by the compliance properties of the heart and and studies dating back over 50 years have shown that in gen- the kinetics of relaxation, especially at higher HRs. Preload is eral, for each 1 ml increase in VO2, there is an approximately affected by venous return, diastolic function, the right heart, 6 ml increase in QC in healthy humans.30,31 and the pericardium. SV increases by approximately 40% during upright exer- The ventricle can be characterized by its capacity to eject cise, owing to both augmentation in LVEDV and reduction in blood (systolic function) and to fill with blood (diastolic func- LV endsystolic volume (ESV).29 Increases in RV EDV occur tion). Systolic function reflects the heart’s pumping ability and in response to enhanced venous return mediated by muscle and is most often measured clinically by the EF. However, EF ventilatory pumps, and possibly dynamic reductions in blood varies directly with preload and inversely with afterload (as is volume in the large-capacity splanchnic veins.27 This increase true of all measures of systolic shortening or thickening). Al- in venous return to the thorax during early exercise is opti- ternative measures that account for both preload and afterload, mally coupled to 2 compliant ventricles that can accommodate such as LV endsystolic elastance (Ees), stress-corrected frac- increases in EDV without excessive rises in ventricular filling tional shortening, or preload-recruitable stroke work are more pressure (end-diastolic pressures, EDP). Net ventricular cham- specific measures of LV chamber contractility that are inde- ber compliance (change in volume per change in pressure) is pendent of load. related to chamber volumes as well to factors within the car- Diastolic function is often dichotomized into “active” and diac myocyte (eg, cytosolic calcium levels, titin phosphoryla- “passive” functions, though these distinctions are arbitrary and tion and isotype expression), the extracellular matrix (collagen there is considerable overlap between them. “Active” function content and cross-linking), the right heart and the pericardi- generally refers to early diastolic processes involved in cross- um.32 Increases in EDV lead to increased stretch of cardiac bridge detachment, calcium reuptake and rapid pressure decay myocytes that serves to enhance the force of contraction via during isovolumic relaxation and early filling. Active diastolic the Frank-Starling mechanism. To allow the ventricle to fill to function is usually measured by the time constant of early dia- a larger EDV in a shorter interval of time, without pathologic stolic pressure decay (τ) or the velocity/extent of tissue motion increases in EDP, there is exercise-induced enhancement in during early filling. Passive diastolic function generally refers early diastolic relaxation, suction and untwist that is mediated to the mechanical properties of heart muscle that dictate the predominantly by adrenergic stimulation.33 Recent research degree of pressure elevation to achieve a given preload vol- indicates that LV chamber stiffness can be dynamically re- ume (end-diastolic chamber elastance or Eed, an approxima- duced by cyclic adenosine monophosphate- (cAMP) and cy- tion for the slope of the end-diastolic pressure-volume rela- clic guanosine monophosphate- (cGMP) dependent phosphor- tionship). Eed is determined by stiffness within the cardiac ylation of macromolecules in the cardiac myocyte, such as myocytes as well as the extracellular matrix, chamber, and titin,34–36 though it is unknown if this protein phosphorylation pericardium. affects elastance (stiffness) acutely during exercise. Protein If systolic and diastolic ventricular properties are held con- kinase C also phosphorylates titin, but in contrast to the stant, an isolated increase in afterload will increase blood pres- cAMP- and cGMP-dependent pathways, PKC phosphoryla- sure and decrease SV, whereas an isolated increase in preload tion increases stiffness.37 Titin affects chamber stiffness via will increase blood pressure and increase SV. The converse additional phosphorylation-independent mechanisms such as effects will occur with reductions in afterload and preload. variable expression ratios between the more complaint (N2BA) Autonomic tone potently influences cardiovascular status: in- and Stiff (N2B) titin isoforms as well oxidation of N2B spring creases in sympathetic stimulation peripherally enhance ve- elements.38 nous return, contractility, chronotropy and lusitropy, whereas To achieve adequate filling of the LV, there must be suffi- increases in parasympathetic tone reduce HR and vascular re- cient augmentation of RV ejection during exercise. This is sistance. Ventricular reserve function reflects the heart’s capac- achieved via RV contractile reserve as well as by pulmonary ity to deal with the increases in preload and afterload that ac- artery (PA) vasodilation.22 Owing to the high compliance and company physiologic stresses such as exercise. As shall be low resistance in the pulmonary vasculature, normal individu- discussed, reserve capacity is often dissonant with resting, als accommodate large increases in pulmonary cardiac output steady state cardiovascular function, particularly in patients (QP) without significant increase in PA pressure. However, with HFpEF. recent studies have suggested that at least mild to moderate increases in PA pressure during exercise may be part of nor- mal aging, particularly above the age of 50.39 Normal Exercise Physiology Although EDV may increase by 20–40%, this reserve be- To cope with the increased metabolic demands of physical ex- comes exhausted quite early on during the course of exercise, ercise, the human body relies upon complex interactions among and further increases in SV must be mediated by more vigorous the heart, lungs, vasculature, endothelium, skeletal muscle emptying of the heart (reduction in ESV).27,29 This is achieved and autonomic nervous system.27,28 Exercise capacity is most by increases in contractility and enhanced arterial vasodilation. often quantified as the peak oxygen consumption (peak VO2) Depending on the parameter used to assess it, LV contractility achieved during maximal effort exercise. According to the Fick increases 1–3-fold during exercise.15 There is also vasorelax- principle, VO2 is defined by the product of cardiac output (QC) ation, mediated by muscular arteriolar dilation, endothelium- and arterial-venous oxygen content difference (AVO2diff). In dependent dilation and reduction in mean SVR.27 However, normal men, VO2 increases 7.7-fold during maximal exercise, total arterial afterload, expressed by effective Ea, increases at achieved by a 3.1-fold increase in QC and 2.5-fold increase in maximal exercise.40 This is because increases in HR and de- AVO2diff.29 Cardiac output is equal to the product of SV and creases in compliance exceed the reduction in SVR at peak HR, whereas AVO2diff is determined by the abilities to oxy- exertion.25 The coupling of the heart to the vasculature can be genate blood in the lungs, transport O2 to tissues bound to expressed by the ratio of Ea to Ees, a load-independent mea- hemoglobin, and then distribute, extract and utilize O2 in ex- sure of contractility.25,40 This coupling ratio (Ea/Ees) varies ercising muscle. The variable that “drives” increases in QC inversely with EF and drops significantly during exercise to

Circulation Journal Vol.78, January 2014 22 BORLAUG BA

Figure 1. Diastolic reserve limitations during exercise in patients with heart failure with preserved ejection fraction (HFpEF). (A) In a typical HFpEF patient, there is dramatic elevation in mean left ventricular diastolic pressure (mLVDP) during exercise associ- ated with limited hastening of isovolumic relaxation (reduction in the time constant of relaxation, τ). Inadequate relaxation reserve during exercise is coupled with an acute shift upward and to the left in the diastolic pressure-volume relationship, with an increase in operant diastolic elastance (Ed) as shown in an example patient (B) and in group data (C,D). The solid lines/bar graphs show raw unadjusted pressure-volume data and the dashed lines/hashed bar graphs show elastance data corrected for the effects of incomplete relaxation. (Adapted with permission from Borlaug et al.47)

allow for greater ejection capacity. blood volume), meaning that without regional vasoconstriction SV recruitment typically plateaus during exertion at ap- during exercise, systemic hypotension would routinely devel- proximately 50% peak VO2, and further increases in QC are op. The capacity of the vasculature to rapidly dilate or constrict achieved purely by HR.29 The rapid increase in HR from rest allows for redistribution of blood volume and flow to meet to approximately 100 beats/min is mediated by acute with- metabolic requirements while optimizing blood pressure to drawal of parasympathetic tone; further increases are mediated maintain perfusion. Many of these changes in local perfusion by increased sympathetic outflow.27 The dominant contributor are regulated by central neural command,27 but distribution of to the age-related drop in maximal QC during exercise is loss flow is also importantly regulated at the tissue level by factors of chronotropic reserve.41 generated by contracting skeletal muscle. Local decreases in Enhanced peripheral O2 distribution, utilization and extrac- pH as well as increases in potassium, adenosine diphosphate, tion (ie, AVO2diff reserve) during exercise plays an equally carbon dioxide, and temperature serve to dilate arterioles and important role as QC reserve.18,27 Although the heart increases precapillary sphincters to optimally couple delivery of blood its output, this enhanced flow needs to be matched to tissues flow to meeting local demands. Among the local factors gener- where perfusion is most needed, which is achieved by regional ated, elaboration of nitric oxide (NO) plays a critically impor- vasodilation in skeletal and cardiac muscle and vasoconstric- tant role in enhancing regional blood flow during stress.42 tion in non-exercising regions such as the skin, splanchnic Chemoreceptors and metaboreceptors located within muscle beds, and kidneys.27 The importance of the latter is appreciated and vasculature provide further neural feedback that modulates by considering that when fully dilated, the body’s blood ves- central efferent autonomic output and may contribute to symp- sels can hold over 20 L of blood (4-fold greater than the usual toms of dyspnea and .43,44

Circulation Journal Vol.78, January 2014 Exercise Intolerance in HFpEF 23

Figure 2. Filling pressures and pulmonary artery (PA) pressures during exercise in early heart failure with preserved ejection fraction (HFpEF). (A) Despite normal resting pulmonary capillary wedge pressures (PCWP), patients with early-stage HFpEF de- velop a dramatic elevation in PCWP with even 1 min of low level (20 Watts) exercise, that rapidly returns to baseline values with cessation of exercise. (B) This increase in PCWP is secondary to exercise-induced elevation in LV end-diastolic pressure (LVEDP) in HFpEF and is coupled with secondary, passive elevation in PA pressures (C), causing exercise-induced PA hypertension. (Adapted with permission from Borlaug et al.49)

function, it does not appear that exercise limitation in HFpEF Pathophysiology of Exercise Intolerance in HFpEF is related to an inability to adequately fill the LV to a large Diastolic Limitations enough EDV. Diastolic dysfunction remains fundamental to HFpEF and is However, despite apparent preservation of EDV reserve, the most richly studied component of the pathophysiology. In there is compelling evidence for failure of the Frank-Starling the first study to examine the hemodynamic changes during mechanism (the ability to translate an increase in filling pres- exercise in HFpEF, Kitzman et al6 used nuclear scintigraphy sure to an increase in cardiac ejection) in HFpEF. Shibata et al to assess LV volumes during right heart catheterization of 7 observed that the increase in SV relative to beat-to-beat chang- HFpEF patients and 10 age-matched controls performing max- es in LV end-diastolic pressure (LVEDP: estimated by PA imal effort upright exercise.6 They observed markedly de- diastolic pressure) was impaired in HFpEF patients compared pressed peak VO2 in the HFpEF patients, which was related to age-matched controls.45 In a separate study, Abudiab et al predominantly to an inability to enhance SV and EDV. Despite noted that for any increase in LVEDP, the increase in cardiac the flat EDV response, patients developed marked increases in output was markedly attenuated in patients with HFpEF com- EDP (estimated as the pulmonary capillary wedge pressure, pared to controls.23 PCWP), leading the authors to conclude that exercise intoler- Clinically, this failure of the Frank-Starling reserve is most ance in HFpEF is related predominantly to failure of the Frank- clearly manifest as an elevation in LVFP, and recent studies Starling mechanism. However, more recent studies have ques- have provided insight into the pathogenesis of this process. tioned whether the increase in EDV with exercise is truly Westermann et al performed conductance catheter analysis in impaired in HFpEF. Borlaug et al performed maximal effort 70 HFpEF patients and 20 controls, and observed both pro- exercise testing with simultaneous nuclear scintigraphy in longed relaxation and increased diastolic elastance.46 Although HFpEF patients compared with closely matched aged hyper- elastance did not change during isometric handgrip, filling tensive controls, and observed no difference in the ability to pressures increased as would be predicted by the upward shift recruit EDV during exercise.10 There was no correlation be- in the resting LV diastolic pressure-volume relationship. In tween the change in EDV with exercise and peak VO2 in that another invasive exercise study, Borlaug et al measured LV study, further questioning the contributory role of reduced pressures using high-fidelity micromanometer catheters with aerobic capacity. A more recent study using echocardiogra- simultaneous echocardiography to determine LV volumes at phy to assess LV volumes noted a slightly greater increase in rest and during dynamic supine exercise in patients with EDV with exercise in HFpEF patients as compared to con- HFpEF.47 LV diastolic pressures increased by approximately trols.18 Thus, despite the common presence of diastolic dys- 50% with exercise (P<0.0001), despite a trend towards a re-

Circulation Journal Vol.78, January 2014 24 BORLAUG BA

Figure 3. Contractile reserve limitations during exercise in heart failure with preserved ejection fraction (HFpEF). Bar graphs show the changes from baseline to 20 Watts of workload in preload-recruitable stroke work (PRSW, A), endsystolic elastance (Ees, B), and peak power index (PWR/EDV) in healthy controls (blue), asymptomatic hypertensive patients (HTN, green) and HFpEF patients (red). Even at low level submaximal workload, there is substantial systolic reserve limitation in the HFpEF patients when using these load-independent measures. (D) The ability to enhance contractility tightly correlates with exercise capacity (peak VO2), support- ing a mechanistic contribution. (Adapted with permission from Borlaug et al.15)

duction in LVEDV, coupled with a 50% increase in end-dia- who are clinically euvolemic, with normal resting hemody- stolic LV elastance (Figure 1). The single-beat diastolic vol- namics and normal B-type natriuretic peptide (BNP) levels.49 ume relationship not only steepens (increased stiffness) but The increase in LVEDP during exercise in HFpEF is extraor- shifts upward in parallel, suggesting some element of pericar- dinarily rapid, occurring in the first minute at low workload, dial restraint that likely contributes to the increase in filling and mercurial, with pressures rapidly returning to normal al- pressures. As such, LV minimal diastolic pressure, which most immediately upon cessation of activity (Figure 2).49,50 normally decreases during exercise to enhance filling by “suc- This intermittent nature of the filling pressure elevation con- tion”, was observed to increase by 55% in patients with founds treatment and also likely explains why BNP levels are HFpEF (Figure 1). LV relaxation, assessed by τ was pro- frequently normal in HFpEF.4 Abnormal patterns of gas ex- longed at rest, and importantly, the normal hastening of relax- change correlate with the kinetics of elevated filling pressures ation (decrease in τ) was impaired in HFpEF, such that the in patients with HFpEF undergoing exercise during cardiac proportion of diastole that elapsed prior to complete relax- catheterization, providing further insight into the nature of ation actually increased as opposed to the normal reduction.47 expired gas abnormalities observed during metabolic stress Exercise-induced increases in LVEDP correlated with chang- testing.51 es in diastolic relaxation rates, chamber stiffness and Ea, but In a noninvasive echocardiographic exercise study, Tan were not related to alterations in EDV, HR or QC. In another et al found that patients with HFpEF displayed lower tissue study, Wachter et al observed that rapid atrial pacing was as- Doppler early diastolic (E’) velocities at rest coupled with sociated with less enhancement in LV relaxation and greater inadequate enhancement in E’ during exercise.14 Similar im- reduction in EDV in patients with HFpEF as compared to pairments in E’ reserve have been reported in response to controls.48 dobutamine infusion in HFpEF.17 In addition to relaxation, E’ Elevated LVFP in HFpEF is observed even among patients velocity is determined by left atrial pressure at the mitral valve

Circulation Journal Vol.78, January 2014 Exercise Intolerance in HFpEF 25

Figure 4. Endothelial dysfunction in heart failure with preserved ejection fraction (HFpEF). (A) The hyperemic increase in digital blood flow in response to upper arm cuff occlusion is impaired in HFpEF patients (red) compared to healthy controls (blue). (B) Nearly half of the patients with HFpEF showed evidence of endothelial dysfunction. Subjects with endothelial dysfunction (ED) reported more severe symptoms of fatigue (C) and dyspnea (E) at low-level, submaximal exercise than subjects with normal en- dothelial function, and the reactive hyperemia index (a measure of endothelial function) inversely correlated with symptom sever- ity during exercise (D,F) and directly correlated with peak VO2 (P<0.0001, not shown). HTN, hypertension. (Adapted with permis- sion from Borlaug et al.15)

opening and the extent/velocity of LV untwisting during early studies have shown that despite overall preservation of EF, diastole.52 Tan et al found that at rest, the extent and velocity patients with HFpEF display subtle but significant abnormali- of LV untwisting was impaired in HFpEF patients compared ties in chamber and myocardial contractility,12 as well as im- with controls, and further tended to be impaired during dy- pairments in regional deformation as detected by tissue Doppler namic exercise, indicating that in addition to inadequate relax- and strain-based imaging techniques.7,8,14,53,54 In addition, al- ation reserve, limitations in LV untwisting contribute.14 Using though the resting EF is normal, the exercise-induced enhance- nuclear scintigraphy, Phan et al observed that while peak dia- ment in EF is impaired in patients with HFpEF when compared stolic filling rates were similar in HFpEF patients and controls to matched controls.10,15,55 Impaired EF reserve could be caused at rest, the ability to enhance the time to peak filling was mark- by afterload mismatch or impaired contractile reserve, and al- edly attenuated in HFpEF.13 Intriguingly, those authors also though there is clearly evidence of vascular stiffening and observed impaired cardiac energetics on nuclear magnetic abnormal vasorelaxation in HFpEF,10,15,55,56 Borlaug et al used resonance spectroscopy, which may contribute to this loss of load-independent measures to show that contractile reserve is diastolic reserve. markedly impaired in HFpEF patients compared to age- matched healthy and hypertensive controls, even at low sub- Systolic Limitations maximal matched workloads (Figure 3).15 In their study, the Although EF is the measure that is used most often clinically magnitude of contractile response directly correlated with to assess systolic function, it is a rather poor measure of LV peak VO2 (Figure 3D), further supporting a role for systolic contractility, because of its high load-dependence.25 Numerous reserve dysfunction in the exercise intolerance in HFpEF.

Circulation Journal Vol.78, January 2014 26 BORLAUG BA

Figure 5. Chronotropic incompetence in heart failure with preserved ejection fraction (HFpEF). (A) Compared with controls (red), patients with HFpEF (black) display markedly lower increases in heart rate (∆HR) during maximal effort exercise. (B) The increase in HR with exercise correlates with exercise capacity (peak VO2) in both groups (bivariate P<0.0001), but for any increase in VO2, there was less enhancement of HR in HFpEF patients compared to controls. This difference persisted after adjusting for age and β-blocker use. Heart rate acceleration during the onset of exercise (C) and deceleration during early recovery (D) were abnormal in HFpEF compared to controls. (Data taken from Borlaug et al.15 and Abudiab et al.23)

Similar impairments in systolic reserve in HFpEF have been viding further evidence of inadequate vasodilation.15 The in- described with dobutamine infusion17,19 and rapid pacing57 in ability to adequately dilate, coupled with the inability to enhance HFpEF patients, though not all studies have observed systolic contractility, leads to dynamic abnormalities in ventricular- limitations.20,48 Even subtle impairment of contractility at rest arterial coupling with exercise in HFpEF.13,15,59 may indicate marked limitations in reserve, which may ex- Central and conduit vessel stiffening is common in HFpEF plain why impaired resting myocardial contractility (despite and contributes to exercise limitation and blood pressure labil- normal EF) predicts increased mortality in HFpEF.12 Systolic ity.59–62 Beyond vascular stiffening, the mechanisms for abnor- reserve limitation in HFpEF affects diastole as well, because mal exercise vasodilation remain unclear. Borlaug et al re- the ability to contract to smaller LV volumes at end-systole ported that HFpEF patients commonly display endothelial enhances the recoil and suction forces during early diastole,52 dysfunction, and the extent of dysfunction directly correlated which are also known to be impaired in HFpEF.14,58 with greater symptoms of dyspnea and fatigue during sub- maximal exercise and inversely correlated with peak VO2 in Vascular Stiffening and Dysfunction HFpEF (Figure 4).15 Recently, Akiyama et al reported that the In addition to impaired contractile reserve, inadequate vasodi- presence of endothelial function predicts increased risk of HF lation appears to contribute to the inability to reduce ESV in hospitalization in patients with HFpEF.63 In an elegant study, HFpEF. Borlaug et al first reported attenuated reductions in Guazzi et al administered the phosphodiesterase inhibitor mean SVR during matched and peak exercise in HFpEF,10 and sildenafil to patients with HFrEF, and noted improvements in most,15,49,50,55,59 though not all18 subsequent studies have cor- endothelial function that closely paralleled decreases in venti- roborated this finding. In another HFpEF population, Borlaug latory responses to stress, supporting the notion that there is et al performed direct assessment of peripheral blood flow cross-talk between microvascular function in skeletal muscle during exercise using a finger-tip plethysmographic device and and symptoms of effort intolerance.44 However, not all studies observed impaired increases in digital flow in HFpEF patients have observed endothelial dysfunction in HFpEF. Haykowsky compared to hypertensive and non-hypertensive controls, pro- et al observed no difference in endothelial function in HFpEF

Circulation Journal Vol.78, January 2014 Exercise Intolerance in HFpEF 27 compared with healthy age-matched controls, or any correla- minants of exercise capacity.76 Intriguingly, the presence of tion between endothelial function and peak VO2 in multivari- AF in HFpEF was found to be an independent predictor of RV ate analysis.64 In another study, this group observed no im- dysfunction, which may additionally explain the adverse out- provements in vascular stiffness or endothelial function after come and greater impairment in exercise capacity in subjects 16 weeks of exercise training in HFpEF.65 This is in contrast with AF.24 However, at this time it remains unclear whether to older studies in non-HF populations where habitual exercise AF causes the RV dysfunction or vice versa. has been associated with improvements in age-associated losses in endothelial function.66 Part of the discrepancy in re- Right Heart, Pulmonary Vasculature and the Pericardium sults between various studies may be related to the methods (PH) is common in HFpEF,77 particu- used to assess endothelial function, which may be performed larly during exercise,49 because downstream elevations in left in larger conduit vessels such as the brachial or femoral arter- heart pressures add in series with components of pressure-re- ies (as in the Haykowsky study) compared with the microvas- lated resistance and QP.22 The presence of PH increases the risk culature (as in the Borlaug and Akiyama studies). Indeed, re- of death in HFpEF,77 and recent studies have shown that pro- cent work from the Framingham group has shown significant gression in pulmonary vascular in HFpEF is associated differences in the risk factor associations and overall preva- with a deterioration in RV function that is potently associated lence of endothelial dysfunction when assessed using these with adverse outcome, independent of other established pre- different methods, suggesting that they may provide distinct dictors of mortality.24 Tedford et al have shown that increases information regarding vascular function.67 in PCWP are associated with lower PA compliance at any Just as systolic dysfunction begets diastolic dysfunction by given resistance, increasing the oscillatory load on the RV, and impairing elastic recoil during early diastole, vascular stiffen- this effect occurs even acutely during exercise in patients with ing may also impair diastolic function.25 Increases in arterial HFpEF.78 Because they are connected in series, inadequate RV afterload prolong relaxation, particularly in the failing heart.68 output could contribute to exercise limitation by promoting Loading sequence appears to also be important, with later- “underfilling” of the left heart during stress. Although this systolic afterload increases having more deleterious effects phenomenon has been demonstrated in HFrEF,79 it has not upon LV ejection and relaxation.69–71 This provides another been documented in patients with HFpEF. mechanism whereby vascular stiffening and elevated afterload Complicating this assessment is the fact that right-sided contribute to elevation in filling pressures observed with stress chamber and left atrial enlargement in HFpEF with RV failure in HFpEF.60 creates the substrate for increased pericardial restraint and dia- stolic ventricular interaction,80 where the right heart influences Heart Rate and Rhythm the left in parallel.22 In this circumstance, left heart filling pres- As discussed earlier, enhancement in HR plays a critical role sures may be elevated even if LVEDV is normal or reduced.81 in allowing for adequate QC reserve, and several studies have Janicki observed that a significant number of patients with identified impairments in chronotropic reserve in HFpEF advanced HFrEF display this phenomenon of enhanced dia- (Figure 5).3,9,10,15,18,23,49,72,73 Borlaug et al observed significant- stolic ventricular interaction during exercise, whereby PCWP ly depressed peak HR in HFpEF patients compared to con- and right atrial pressure increase in tandem with no further trols, which was coupled with abnormalities in cardioaccel- enhancement in SV, because of the restraining effects of the eration and HR recovery after cessation of exercise.10 The pericardium that prevent additional preload recruitment.82 Fur- latter is a marker of autonomic health, which is associated with ther study is required to determine whether this phenomenon longevity, even after controlling for the presence or absence of contributes to exercise intolerance in HFpEF, or perhaps cardiovascular diseases. Other groups have also observed im- whether surgical mitigation of pericardial restraint might im- paired HR recovery in HFpEF, even when accounting for prove exercise capacity. β-blocker usage,73 and abnormal arterial baroreflex sensitivity has also been documented,10 further supporting a role for au- Peripheral Factors tonomic dysfunction in HFpEF. Although not all studies have There are numerous lines of evidence to support the notion observed abnormal HR reserve in HFpEF,50 it has been found that true maximal VO2 is limited by central (QC) rather than to be extremely common in several large recent studies, ranging peripheral (AVO2diff) determinants. These include the obser- from as high as 56% to 78%.15,72 This observation, coupled with vations that (1) VO2 max is achieved by activation of only 50% the tight correlation between HR reserve and peak VO2,9,10,15 of total body muscle mass, and activation of additional muscle suggests that restoration of chronotropic competence may en- does not increase QC or VO2 beyond this level, (2) when ad- hance exercise capacity in HFpEF. However, increases in HR ditional muscles are activated at VO2 max, arterial blood pres- may compromise LV filling in patients with prolonged relax- sure is maintained by vasoconstriction to exercising muscles, ation, and a recent small study reported dramatic improve- and (3) the capacity of skeletal muscle to consume O2 greatly ments in peak VO2 after just 7 days of treatment with the nega- exceeds what the heart can deliver.27 However, true VO2 max is tive chronotropic drug, ivabradine.74 Clearly, further research rarely achieved in patients with HF, and VO2 max is not equiva- is needed regarding the management of HR responses to exer- lent to what is routinely measured in practice (peak VO2). cise in HFpEF patients, and a prospective trial testing the ef- Recently, Haykowsky et al examined the determinants of fects of rate-adaptive pacing in patients with chronotropic in- peak VO2 in patients with HFpEF and age-matched controls competence is currently being planned. free of HF. Although they observed that peak VO2 was tightly In addition to HR reserve, heart rhythm appears to be im- correlated with both QC and AVO2diff, in both their HFpEF portant in HFpEF. A recent study has shown the presence of and control groups, the change in AVO2diff with exercise was (AF) portends greater risk of death in HFpEF the strongest independent predictor of peak VO2.18 In a sepa- patients.75 In an ancillary study from the RELAX trial, Zakeri rate study, this group examined body composition in the same et al demonstrated that despite similar HR responses to exer- cohort and observed a decrease in the percent of lean total body cise, HFpEF patients with AF displayed much more severe and leg mass.83 Although peak VO2 increased with increasing exercise intolerance, even after adjusting for other key deter- lean mass in healthy controls, this was not observed in HFpEF

Circulation Journal Vol.78, January 2014 28 BORLAUG BA

Figure 6. Central vs. peripheral mechanisms of exercise impairment in heart failure with preserved ejection fraction (HFpEF). (A) Compared to controls (red), patients with HFpEF (black) display less increase in cardiac output (∆CO) during exercise. (B) This cardiac limitation is maintained when normalizing ∆CO to metabolic demands (∆VO2). (C) Total increases in arterial-venous O2 content differences (∆AVO2diff) with maximal exercise were similar in HFpEF patients and controls, but when normalized to ∆VO2, there was greater reliance in the HFpEF subjects on enhanced peripheral O2 extraction and utilization (D), likely to compensate for inadequate skeletal muscle perfusion because of CO limitation. (Adapted with permission from Abudiab et al.23)

patients, suggesting that impaired skeletal muscle perfusion or study, for any increase in VO2, there was on average a 20% oxidative capacity might be present. An alternative possibility lower increase in QC in the HFpEF patients (Figure 6), indi- might be that increased QC during exercise is inappropriately cating that limited cardiac output reserve, rather than periph- directed to non-exercising tissues, such as adipose.83 In a dif- eral dysfunction, was the dominant contributor to the reduc- ferent cohort, Bhella et al found that while QC reserve was tion in peak VO2. When contemplating the reasons for the depressed in HFpEF (as previously demonstrated by other divergent findings between this study and those of Haykowsky groups), when normalized to VO2, HFpEF patients displayed and Bhella, it is important to consider that HFpEF is a hetero- an overly exuberant increase in QC, with markedly impaired geneous disease, and it is highly likely that there are subgroups AVO2diff reserve.20 Those authors proposed the provocative of patients in whom exercise capacity is more or less con- conclusion that exercise capacity in HFpEF is limited by pre- strained by central vs. peripheral mechanisms. Further efforts mature skeletal muscle fatigue and/or by metabolic/neural sig- to subtype these categories may be helpful in the classification nals originating in muscle that stimulate excessive QC respons- and individualization of treatment of HFpEF.23,84 es to exercise. Regardless of the primacy of peripheral and skeletal muscle However, in a subsequently published, invasive study in abnormalities in limiting peak VO2 in HFpEF, they could be which arterial and mixed venous O2 contents were directly highly viable therapeutic targets.85 Haykowsky et al noted that measured, no differences in maximal exercise AVO2diff were improved peak VO2 in HFpEF patients after 4 months of ex- observed between HFpEF patients and controls (Figure 6).23 ercise training was predominantly related to higher AVO2diff Furthermore, when AVO2diff was normalized to peak VO2 at peak exercise, with no significant change in CQ .86 It may be achieved, it was higher rather than lower in patients with that there is greater plasticity and potential for rejuvenation in HFpEF. The authors speculated that this represents a chronic peripheral factors such as blood flow distribution, microvas- adaptation to inadequate QC, citing the fact that with a lower cular function, capillary density, skeletal muscle function, QC, there is greater time available for gas exchange in the energetic stores and mitochondrial oxidative capacity as com- capillaries, which may serve to counterbalance any potential pared with cardiac properties.87 Although some groups have abnormalities in convective and diffusive O2 transfer.23 In that noted improvements in the LV diastolic function of HFpEF

Circulation Journal Vol.78, January 2014 Exercise Intolerance in HFpEF 29

Figure 7. Summary of mechanisms of exercise intolerance in heart failure with preserved ejection fraction (HFpEF). “Central” impairments are shown in the ovals and “peripheral” impairments in the rectangles. Arrows indicate where an abnormality di- rectly contributes to another. Question marks indicate theoretical but unproven mechanistic relationships. See text for additional details. AVO2, arterial-venous oxygen content; β-AR, beta-adrenoreceptor; LVFP, left ventricular filling pressures; PA, pulmonary artery; RV, right ventricle; VA, ventricular-arterial.

patients doing exercise training,88 a more recent study ob- therapies failed to detect an improvement in submaximal ex- served no changes in cardiovascular mechanics.89 Although ercise capacity (6-min walk distance) of HFpEF patients, de- the latter study (n=11) reported no change in peak VO2 with spite a significant increase in hemoglobin while on treatment.90 exercise training, 2 much larger randomized trials (n=64 and Figure 7 summarizes the roles of multiple mechanisms in the 46, respectively) have demonstrated fairly substantial im- pathogenesis of HFpEF as has been described. provements in aerobic capacity with training (2.6–3.3 ml · min–1 · kg–1).87,88 Exercise Reserve, HFpEF and Normal Aging Other Factors LV diastolic stiffness increases with normal aging,91 and the The lungs function as the first step in the transport of 2O from age-related increase is greater in women than in men and cor- the external environment to the tissues for utilization, and ab- relates with increases in body mass, all consistent with the normalities in ventilation or gas exchange (alveolar membrane typical demographic characteristics of HFpEF: older, hyperten- O2 conductance, capillary blood volume, ventilation-perfusion sive, obese females.92,93 In addition, the ability to tolerate vol- matching) might be expected to contribute to exercise impair- ume loading without excessive increase in LVFP decreases in ment in HF. Pulmonary limitations have not been considered older women.94 Given these observations, together with the as a significant contributor to reduced peak VO2, given the foregoing findings, therapies targeting age-related LV diastolic dramatic reserve capacity of the lungs,27 but pulmonary me- stiffening would appear to hold promise to help improve exer- chanics and gas exchange have not been adequately studied cise capacity and outcomes in HFpEF.95 Indeed, better under- in HFpEF, and it is certainly conceivable that with acute in- standing of the reasons for age-related reductions in cardiac creases in LVFP, interstitial edema develops, changing the reserve96 and exercise capacity97 may allow for the develop- compliance properties of the lungs, possibly leading to greater ment of novel therapies to improve outcome in both HFpEF respiratory muscle fatigue that may discourage further exer- patients and the broader population of older adults without HF cise. Adequate delivery of O2 from the lungs to muscle and but with deterioration in functional capacity related to normal tissues requires hemoglobin, and anemia is common in HFpEF aging. patients. However, a recent trial of erythropoietin-stimulating

Circulation Journal Vol.78, January 2014 30 BORLAUG BA

10. Borlaug BA, Melenovsky V, Russell SD, Kessler K, Pacak K, Conclusions and Future Directions Becker LC, et al. Impaired chronotropic and vasodilator reserves In summary, numerous recent studies have changed the way limit exercise capacity in patients with heart failure and a preserved ejection fraction. Circulation 2006; 114: 2138 – 2147. we conceptualize the pathophysiology of HFpEF: away from 11. Lam CS, Roger VL, Rodeheffer RJ, Bursi F, Borlaug BA, Ommen being simply a disorder of diastolic dysfunction to more of a SR, et al. Cardiac structure and ventricular-vascular function in per- global impairment in the multiple components of cardiovascu- sons with heart failure and preserved ejection fraction from Olm- lar15,23 and peripheral18,20 reserves, which limits the ability to sted County, Minnesota. Circulation 2007; 115: 1982 – 1990. 12. Borlaug BA, Lam CS, Roger VL, Rodeheffer RJ, Redfield MM. cope with physiologic stresses such as exercise. Further re- Contractility and ventricular systolic stiffening in hypertensive heart search is required to refine our understanding, enabling more disease insights into the pathogenesis of heart failure with preserved optimal treatment for patients with HFpEF. A potential area of ejection fraction. J Am Coll Cardiol 2009; 54: 410 – 418. particular interest might be rigorous phenotyping of the nature 13. Phan TT, Abozguia K, Nallur Shivu G, Mahadevan G, Ahmed I, Williams L, et al. Heart failure with preserved ejection fraction is of exercise limitation in a given patient (eg, predominant dia- characterized by dynamic impairment of active relaxation and con- stolic vs. systolic limitation, or peripheral vs. central limita- traction of the left ventricle on exercise and associated with myocar- tion), which may allow for more precise individualization of dial energy deficiency. J Am Coll Cardiol 2009; 54: 402 – 409. therapies.23,84 Clearly, better understanding the mechanisms of 14. Tan YT, Wenzelburger F, Lee E, Heatlie G, Leyva F, Patel K, et al. The pathophysiology of heart failure with normal ejection fraction: the age-related reductions in cardiovascular reserve from the Exercise echocardiography reveals complex abnormalities of both cellular to the organ system level needs further probing, and systolic and diastolic ventricular function involving torsion, untwist, innovative new studies have just begun to make strides in this and longitudinal motion. J Am Coll Cardiol 2009; 54: 36 – 46. regard.98 Given the plurality of reserve limitations noted in 15. Borlaug BA, Olson TP, Lam CS, Flood KS, Lerman A, Johnson BD, HFpEF, it begs the question as to whether there are overarch- et al. Global cardiovascular reserve dysfunction in heart failure with preserved ejection fraction. J Am Coll Cardiol 2010; 56: 845 – 854. ing or unifying processes that affect the ventricle, vasculature 16. Prasad A, Hastings JL, Shibata S, Popovic ZB, Arbab-Zadeh A, and periphery that can be treated to improve global reserve, Bhella PS, et al. Characterization of static and dynamic left ventricu- with possibilities including increased nitroso-oxidative stress, lar diastolic function in patients with heart failure with a preserved abnormal NO/cGMP-dependent signaling pathways,99 or tis- ejection fraction. Circ Heart Fail 2010; 3: 617 – 626. 17. Lee AP, Song JK, Yip GW, Zhang Q, Zhu TG, Li C, et al. Impor- sue deposition of misfolded proteins acquired with senes- tance of dynamic dyssynchrony in the occurrence of hypertensive cence.100 Finally, although clinical trials in HF have tradition- heart failure with normal ejection fraction. Eur Heart J 2010; 31: ally targeted mortality and HF hospitalizations, it is time to 2642 – 2649. think more globally and remember that quality of life is often 18. Haykowsky MJ, Brubaker PH, John JM, Stewart KP, Morgan TM, Kitzman DW. Determinants of exercise intolerance in elderly heart just as important as quantity in patients with HFpEF. It is an failure patients with preserved ejection fraction. J Am Coll Cardiol inescapable fact that the number of people with HFpEF is 2011; 58: 265 – 274. going to increase dramatically during the next 30 years, and 19. Norman HS, Oujiri J, Larue SJ, Chapman CB, Margulies KB, further studies examining the determinants and opportunities Sweitzer NK. Decreased cardiac functional reserve in heart failure with preserved systolic function. J Card Fail 2011; 17: 301 – 308. for improvement of exercise intolerance are critically impor- 20. Bhella PS, Prasad A, Heinicke K, Hastings JL, Arbab-Zadeh A, tant to ensure that these people can enjoy independent, produc- Adams-Huet B, et al. Abnormal haemodynamic response to exer- tive and active lifestyles in the years to come. cise in heart failure with preserved ejection fraction. Eur J Heart Fail 2011; 13: 1296 – 1304. 21. Guazzi M, Vicenzi M, Arena R, Guazzi MD. Pulmonary hyperten- Disclosures sion in heart failure with preserved ejection fraction: A target of Conflicts: None. phosphodiesterase-5 inhibition in a 1-year study. Circulation 2011; 124: 164 – 174. 22. Guazzi M, Borlaug BA. Pulmonary hypertension due to left heart References disease. Circulation 2012; 126: 975 – 990. 1. Sakata Y, Shimokawa H. Epidemiology of heart failure in asia. 23. Abudiab MM, Redfield MM, Melenovsky V, Olson TP, Kass DA, Circ J 2013; 77: 2209 – 2217. Johnson BD, et al. Cardiac output response to exercise in relation 2. Owan TE, Hodge DO, Herges RM, Jacobsen SJ, Roger VL, Redfield to metabolic demand in heart failure with preserved ejection frac- MM. Trends in prevalence and outcome of heart failure with pre- tion. Eur J Heart Fail 2013; 15: 776 – 785. served ejection fraction. N Engl J Med 2006; 355: 251 – 259. 24. Melenovsky V, Hwang S, Lin G, Redfield MM, Borlaug BA. Right 3. Kitzman DW, Little WC, Brubaker PH, Anderson RT, Hundley heart dysfunction in heart failure with preserved ejection fraction: WG, Marburger CT, et al. Pathophysiological characterization of The determinants and prognostic impact. Circulation 2013 [Abstract]. isolated diastolic heart failure in comparison to systolic heart fail- 25. Borlaug BA, Kass DA. Ventricular-vascular interaction in heart ure. JAMA 2002; 288: 2144 – 2150. failure. Cardiol Clin 2011; 29: 447 – 459. 4. Borlaug BA, Paulus WJ. Heart failure with preserved ejection frac- 26. Borlaug BA, Kass DA. Invasive hemodynamic assessment in heart tion: Pathophysiology, diagnosis, and treatment. Eur Heart J 2011; failure. Cardiol Clin 2011; 29: 269 – 280. 32: 670 – 679. 27. Rowell LB. Human cardiovascular control. Oxford: Oxford Univer- 5. Zile MR, Baicu CF, Gaasch WH. Diastolic heart failure: Abnor- sity Press, 1993. malities in active relaxation and passive stiffness of the left ventri- 28. Kitzman DW, Groban L. Exercise intolerance. Cardiol Clin 2011; cle. N Engl J Med 2004; 350: 1953 – 1959. 29: 461 – 477. 6. Kitzman DW, Higginbotham MB, Cobb FR, Sheikh KH, Sullivan 29. Higginbotham MB, Morris KG, Williams RS, McHale PA, Coleman MJ. Exercise intolerance in patients with heart failure and preserved RE, Cobb FR. Regulation of stroke volume during submaximal and left ventricular systolic function: Failure of the Frank-Starling mech- maximal upright exercise in normal man. Circ Res 1986; 58: 281 – anism. J Am Coll Cardiol 1991; 17: 1065 – 1072. 291. 7. Yu CM, Lin H, Yang H, Kong SL, Zhang Q, Lee SW. Progression 30. Astrand PO, Cuddy TE, Saltin B, Stenberg J. Cardiac output during of systolic abnormalities in patients with “isolated” diastolic heart submaximal and maximal work. J Appl Physiol 1964; 19: 268 – 274. failure and diastolic dysfunction. Circulation 2002; 105: 1195 – 1201. 31. Beck KC, Randolph LN, Bailey KR, Wood CM, Snyder EM, 8. Brucks S, Little WC, Chao T, Kitzman DW, Wesley-Farrington D, Johnson BD. Relationship between cardiac output and oxygen con- Gandhi S, et al. Contribution of left ventricular diastolic dysfunc- sumption during upright cycle exercise in healthy humans. J Appl tion to heart failure regardless of ejection fraction. Am J Cardiol Physiol 2006; 101: 1474 – 1480. 2005; 95: 603 – 606. 32. Borlaug BA, Kass DA. Mechanisms of diastolic dysfunction in heart 9. Brubaker PH, Joo KC, Stewart KP, Fray B, Moore B, Kitzman DW. failure. Trends Cardiovasc Med 2006; 16: 273 – 279. Chronotropic incompetence and its contribution to exercise intoler- 33. Cheng CP, Igarashi Y, Little WC. Mechanism of augmented rate of ance in older heart failure patients. J Cardiopulm Rehabil 2006; 26: left ventricular filling during exercise. Circ Res 1992; 70: 9 – 19. 86 – 89. 34. van Heerebeek L, Hamdani N, Falcao-Pires I, Leite-Moreira AF,

Circulation Journal Vol.78, January 2014 Exercise Intolerance in HFpEF 31

Begieneman MP, Bronzwaer JG, et al. Low myocardial protein Cardiol 2001; 38: 796 – 802. kinase g activity in heart failure with preserved ejection fraction. 57. Liu CP, Ting CT, Lawrence W, Maughan WL, Chang MS, Kass Circulation 2012; 126: 830 – 839. DA. Diminished contractile response to increased heart rate in intact 35. Borbely A, Falcao-Pires I, van Heerebeek L, Hamdani N, Edes I, human left ventricular hypertrophy: Systolic versus diastolic deter- Gavina C, et al. Hypophosphorylation of the Stiff N2B titin isoform minants. Circulation 1993; 88: 1893 – 1906. raises cardiomyocyte resting tension in failing human myocardium. 58. Ohara T, Niebel CL, Stewart KC, Charonko JJ, Pu M, Vlachos PP, Circ Res 2009; 104: 780 – 786. et al. Loss of adrenergic augmentation of diastolic intra-lv pressure 36. van Heerebeek L, Borbely A, Niessen HW, Bronzwaer JG, van der difference in patients with diastolic dysfunction: Evaluation by color Velden J, Stienen GJ, et al. Myocardial structure and function differ M-mode echocardiography. JACC Cardiovasc Imaging 2012; 5: in systolic and diastolic heart failure. Circulation 2006; 113: 1966 – 861 – 870. 1973. 59. Tartiere-Kesri L, Tartiere JM, Logeart D, Beauvais F, Cohen Solal 37. Hidalgo C, Hudson B, Bogomolovas J, Zhu Y, Anderson B, Greaser A. Increased proximal arterial stiffness and cardiac response with M, et al. PKC phosphorylation of titin’s PEVK element: A novel moderate exercise in patients with heart failure and preserved ejec- and conserved pathway for modulating myocardial stiffness. Circ tion fraction. J Am Coll Cardiol 2012; 59: 455 – 461. Res 2009; 105: 631 – 638, 17 p following 638. 60. Kawaguchi M, Hay I, Fetics B, Kass DA. Combined ventricular 38. Hidalgo C, Granzier H. Tuning the molecular giant titin through systolic and arterial stiffening in patients with heart failure and pre- phosphorylation: Role in health and disease. Trends Cardiovasc served ejection fraction: Implications for systolic and diastolic re- Med 2013; 23: 165 – 171. serve limitations. Circulation 2003; 107: 714 – 720. 39. Kovacs G, Berghold A, Scheidl S, Olschewski H. Pulmonary arte- 61. Kitzman DW, Herrington DM, Brubaker PH, Moore JB, Eggebeen rial pressure during rest and exercise in healthy subjects: A system- J, Haykowsky MJ. Carotid arterial stiffness and its relationship to atic review. Eur Respir J 2009; 34: 888 – 894. exercise intolerance in older patients with heart failure and preserved 40. Chantler PD, Lakatta EG, Najjar SS. Arterial-ventricular coupling: ejection fraction. Hypertension 2013; 61: 112 – 119. Mechanistic insights into cardiovascular performance at rest and 62. Desai AS, Mitchell GF, Fang JC, Creager MA. Central aortic stiff- during exercise. J Appl Physiol 2008; 105: 1342 – 1351. ness is increased in patients with heart failure and preserved ejec- 41. Brubaker PH, Kitzman DW. Chronotropic incompetence: Causes, tion fraction. J Card Fail 2009; 15: 658 – 664. consequences, and management. Circulation 2011; 123: 1010 – 1020. 63. Akiyama E, Sugiyama S, Matsuzawa Y, Konishi M, Suzuki H, 42. Richardson RS, Saltin B. Human muscle blood flow and metabo- Nozaki T, et al. Incremental prognostic significance of peripheral lism studied in the isolated quadriceps muscles. Med Sci Sports endothelial dysfunction in patients with heart failure with normal Exerc 1998; 30: 28 – 33. left ventricular ejection fraction. J Am Coll Cardiol 2012; 60: 1778 – 43. Middlekauff HR. Making the case for skeletal myopathy as the 1786. major limitation of exercise capacity in heart failure. Circ Heart 64. Haykowsky MJ, Herrington DM, Brubaker PH, Morgan TM, Hundley Fail 2010; 3: 537 – 546. WG, Kitzman DW. Relationship of flow-mediated arterial dilation 44. Guazzi M, Samaja M, Arena R, Vicenzi M, Guazzi MD. Long-term and exercise capacity in older patients with heart failure and pre- use of sildenafil in the therapeutic management of heart failure. J served ejection fraction. J Gerontol A Biol Sci Med Sci 2013; 68: Am Coll Cardiol 2007; 50: 2136 – 2144. 161 – 167. 45. Shibata S, Hastings JL, Prasad A, Fu Q, Bhella PS, Pacini E, et al. 65. Kitzman DW, Brubaker PH, Herrington DM, Morgan TM, Stewart Congestive heart failure with preserved ejection fraction is associ- KP, Hundley WG, et al. Effect of endurance exercise training on ated with severely impaired dynamic starling mechanism. J Appl endothelial function and arterial stiffness in older patients with heart Physiol 2011; 110: 964 – 971. failure and preserved ejection fraction: A randomized, controlled, 46. Westermann D, Kasner M, Steendijk P, Spillmann F, Riad A, single-blind trial. J Am Coll Cardiol 2013; 62: 584 – 592. Weitmann K, et al. Role of left ventricular stiffness in heart failure 66. Golbidi S, Laher I. Exercise and the aging endothelium. J Diabetes with normal ejection fraction. Circulation 2008; 117: 2051 – 2060. Res 2013; 2013: 789607. 47. Borlaug BA, Jaber WA, Ommen SR, Lam CS, Redfield MM, 67. Hamburg NM, Palmisano J, Larson MG, Sullivan LM, Lehman BT, Nishimura RA. Diastolic relaxation and compliance reserve during Vasan RS, et al. Relation of brachial and digital measures of vascu- dynamic exercise in heart failure with preserved ejection fraction. lar function in the community: The Framingham Heart Study. Hy- Heart 2011; 97: 964 – 969. pertension 2011; 57: 390 – 396. 48. Wachter R, Schmidt-Schweda S, Westermann D, Post H, Edelmann 68. Leite-Moreira AF, Lourenco AP, Roncon-Albuquerque R Jr, F, Kasner M, et al. Blunted frequency-dependent upregulation of Henriques-Coelho T, Amorim MJ, Almeida J, et al. Diastolic toler- cardiac output is related to impaired relaxation in diastolic heart ance to systolic pressures closely reflects systolic performance in failure. Eur Heart J 2009; 30: 3027 – 3036. patients with coronary heart disease. Basic Res Cardiol 2012; 107: 49. Borlaug BA, Nishimura RA, Sorajja P, Lam CS, Redfield MM. 251. Exercise hemodynamics enhance diagnosis of early heart failure 69. Borlaug BA, Melenovsky V, Redfield MM, Kessler K, Chang HJ, with preserved ejection fraction. Circ Heart Fail 2010; 3: 588 – 595. Abraham TP, et al. Impact of arterial load and loading sequence on 50. Maeder MT, Thompson BR, Brunner-La Rocca HP, Kaye DM. left ventricular tissue velocities in humans. J Am Coll Cardiol 2007; Hemodynamic basis of exercise limitation in patients with heart 50: 1570 – 1577. failure and normal ejection fraction. J Am Coll Cardiol 2010; 56: 70. Goto T, Ohte N, Fukuta H, Wakami K, Tani T, Kimura G. Relation- 855 – 863. ship between effective arterial elastance, total vascular resistance, 51. Maeder MT, Thompson BR, Htun N, Kaye DM. Hemodynamic and augmentation index at the ascending aorta and left ventricular determinants of the abnormal cardiopulmonary exercise response diastolic function in older women. Circ J 2013; 77: 123 – 129. in heart failure with preserved left ventricular ejection fraction. J 71. Chirinos JA, Segers P, Rietzschel ER, De Buyzere ML, Raja MW, Card Fail 2012; 18: 702 – 710. Claessens T, et al. Early and late systolic wall stress differentially 52. Opdahl A, Remme EW, Helle-Valle T, Lyseggen E, Vartdal T, relate to myocardial contraction and relaxation in middle-aged Pettersen E, et al. Determinants of left ventricular early-diastolic adults: The Asklepios Study. Hypertension 2013; 61: 296 – 303. lengthening velocity: Independent contributions from left ventricu- 72. Redfield MM, Chen HH, Borlaug BA, Semigran MJ, Lee KL, Lewis lar relaxation, restoring forces, and lengthening load. Circulation G, et al. Effect of phosphodiesterase-5 inhibition on exercise capac- 2009; 119: 2578 – 2586. ity and clinical status in heart failure with preserved ejection frac- 53. Wang J, Khoury DS, Yue Y, Torre-Amione G, Nagueh SF. Pre- tion: A randomized clinical trial. JAMA 2013; 309: 1268 – 1277. served left ventricular twist and circumferential deformation, but 73. Phan TT, Nallur Shivu G, Abozguia K, Davies C, Nassimizadeh M, depressed longitudinal and radial deformation in patients with dia- Jimenez D, et al. Impaired heart rate recovery and chronotropic stolic heart failure. Eur Heart J 2008; 29: 1283 – 1289. incompetence in patients with heart failure with preserved ejection 54. Shah AM, Solomon SD. Myocardial deformation imaging: Current fraction. Circ Heart Fail 2009; 3: 29 – 34. status and future directions. Circulation 2012; 125: e244 – e248. 74. Kosmala W, Holland DJ, Rojek A, Wright L, Przewlocka-Kosmala 55. Ennezat PV, Lefetz Y, Marechaux S, Six-Carpentier M, Deklunder M, Marwick TH. Effect of if-channel inhibition on hemodynamics G, Montaigne D, et al. Left ventricular abnormal response during and exercise tolerance in heart failure with preserved ejection frac- dynamic exercise in patients with heart failure and preserved left tion: A randomized trial. J Am Coll Cardiol 2013; 62: 1330 – 1338. ventricular ejection fraction at rest. J Card Fail 2008; 14: 475 – 480. 75. Zakeri R, Chamberlain AM, Roger VL, Redfield MM. Temporal 56. Hundley WG, Kitzman DW, Morgan TM, Hamilton CA, Darty SN, relationship and prognostic significance of atrial fibrillation in heart Stewart KP, et al. Cardiac cycle-dependent changes in aortic area failure patients with preserved ejection fraction: A community- and distensibility are reduced in older patients with isolated dia- based study. Circulation 2013; 128: 1085 – 1093. stolic heart failure and correlate with exercise intolerance. J Am Coll 76. Zakeri R, Borlaug BA, McNulty SE, Mohammed SF, Lewis GD,

Circulation Journal Vol.78, January 2014 32 BORLAUG BA

Semigran MJ, et al. Impact of atrial fibrillation on exercise capac- Stahrenberg R, et al. Exercise training improves exercise capacity ity in heart failure with preserved ejection fraction: A relax trial and diastolic function in patients with heart failure with preserved ancillary study. Circ Heart Fail 2013 October 25, doi:10.1161/ ejection fraction: Results of the EX-DHF (Exercise training in Dia- CIRCHEARTFAILURE.113.000568 [Epub ahead of print]. stolic Heart Failure) pilot study. J Am Coll Cardiol 2011; 58: 1780 – 77. Lam CS, Roger VL, Rodeheffer RJ, Borlaug BA, Enders FT, 1791. Redfield MM. Pulmonary hypertension in heart failure with pre- 89. Fujimoto N, Prasad A, Hastings JL, Bhella PS, Shibata S, Palmer served ejection fraction: A community-based study. J Am Coll Car- D, et al. Cardiovascular effects of 1 year of progressive endurance diol 2009; 53: 1119 – 1126. exercise training in patients with heart failure with preserved ejec- 78. Tedford RJ, Hassoun PM, Mathai SC, Girgis RE, Russell SD, tion fraction. Am Heart J 2012; 164: 869 – 877. Thiemann DR, et al. Pulmonary capillary wedge pressure augments 90. Maurer MS, Teruya S, Chakraborty B, Helmke S, Mancini D. Treat- right ventricular pulsatile loading. Circulation 2012; 125: 289 – 297. ing anemia in older adults with heart failure with a preserved ejec- 79. Butler J, Chomsky DB, Wilson JR. Pulmonary hypertension and tion fraction with epoetin alfa: Single-blind randomized clinical trial exercise intolerance in patients with heart failure. J Am Coll Car- of safety and efficacy. Circ Heart Fail 2013; 6: 254 – 263. diol 1999; 34: 1802 – 1806. 91. Arbab-Zadeh A, Dijk E, Prasad A, Fu Q, Torres P, Zhang R, et al. 80. Melenovsky V, Borlaug BA, Rosen B, Hay I, Ferruci L, Morell CH, Effect of aging and physical activity on left ventricular compliance. et al. Cardiovascular features of heart failure with preserved ejec- Circulation 2004; 110: 1799 – 1805. tion fraction versus nonfailing hypertensive left ventricular hyper- 92. Borlaug BA, Redfield MM, Melenovsky V, Kane GC, Karon BL, trophy in the urban baltimore community: The role of atrial remod- Jacobsen SJ, et al. Longitudinal changes in left ventricular stiffness: eling/dysfunction. J Am Coll Cardiol 2007; 49: 198 – 207. A community-based study. Circ Heart Fail 2013; 6: 944 – 952. 81. Atherton JJ, Moore TD, Lele SS, Thomson HL, Galbraith AJ, 93. Scantlebury DC, Borlaug BA. Why are women more likely than Belenkie I, et al. Diastolic ventricular interaction in chronic heart men to develop heart failure with preserved ejection fraction? Curr failure. Lancet 1997; 349: 1720 – 1724. Opin Cardiol 2011; 26: 562 – 568. 82. Janicki JS. Influence of the pericardium and ventricular interdepen- 94. Fujimoto N, Borlaug BA, Lewis GD, Hastings JL, Shafer KM, dence on left ventricular diastolic and systolic function in patients Bhella PS, et al. Hemodynamic responses to rapid saline loading: with heart failure. Circulation 1990; 81: III15 – III20. The impact of age, sex, and heart failure. Circulation 2013; 127: 83. Haykowsky MJ, Brubaker PH, Morgan TM, Kritchevsky S, Eggebeen 55 – 62. J, Kitzman DW. Impaired aerobic capacity and physical functional 95. Sakata Y, Ohtani T, Takeda Y, Yamamoto K, Mano T. Left ven- performance in older heart failure patients with preserved ejection tricular stiffening as therapeutic target for heart failure with pre- fraction: Role of lean body mass. J Gerontol A Biol Sci Med Sci served ejection fraction. Circ J 2013; 77: 886 – 892. 2013; 68: 968 – 975. 96. Stratton JR, Levy WC, Cerqueira MD, Schwartz RS, Abrass IB. 84. Borlaug BA. Heart failure with preserved and reduced ejection Cardiovascular responses to exercise: Effects of aging and exercise fraction: Different risk profiles for different diseases. Eur Heart J training in healthy men. Circulation 1994; 89: 1648 – 1655. 2013; 34: 1393 – 1395. 97. Fleg JL, Morrell CH, Bos AG, Brant LJ, Talbot LA, Wright JG, et 85. Okita K, Kinugawa S, Tsutsui H. Exercise intolerance in chronic al. Accelerated longitudinal decline of aerobic capacity in healthy heart failure: Skeletal muscle dysfunction and potential therapies. older adults. Circulation 2005; 112: 674 – 682. Circ J 2013; 77: 293 – 300. 98. Loffredo FS, Steinhauser ML, Jay SM, Gannon J, Pancoast JR, 86. Haykowsky MJ, Brubaker PH, Stewart KP, Morgan TM, Eggebeen Yalamanchi P, et al. Growth differentiation factor 11 is a circulat- J, Kitzman DW. Effect of endurance training on the determinants ing factor that reverses age-related cardiac hypertrophy. Cell 2013; of peak exercise oxygen consumption in elderly patients with stable 153: 828 – 839. compensated heart failure and preserved ejection fraction. J Am 99. Paulus WJ, Tschope C. A novel paradigm for heart failure with Coll Cardiol 2012; 60: 120 – 128. preserved ejection fraction: Comorbidities drive myocardial dys- 87. Kitzman DW, Brubaker PH, Morgan TM, Stewart KP, Little WC. function and remodeling through coronary microvascular endothe- Exercise training in older patients with heart failure and preserved lial inflammation. J Am Coll Cardiol 2013; 62: 263 – 271. ejection fraction: A randomized, controlled, single-blind trial. Cir- 100. Willis MS, Patterson C. Proteotoxicity and cardiac dysfunction: culation 2010; 3: 659 – 667. Alzheimer’s disease of the heart? N Engl J Med 2013; 368: 455 – 88. Edelmann F, Gelbrich G, Dungen HD, Frohling S, Wachter R, 464.

Circulation Journal Vol.78, January 2014