INTRODUCTION
The mammalian cardiovascular system is designed to provide adequate flow at physiologic pressures both at rest and under a broad range of demands. Since blood flow is pulsatile, changes in cardiac output are accompanied by alterations in the arterial pulse wave amplitude and peak systolic pressure. To prevent wide fluctuations in blood pressure that otherwise can lead to vascular and end-organ damage, the heart and arteries are compliant so that pulse and peak pressures can be buffered, while systemic diastolic pressures are augmented. For the vasculature, this compliance is largely contained within the proximal conduit vessels, while within the heart, it is described by the end systolic stiffness (elastance, inverse of compliance) that is achieved during contraction. The normal human heart develops a ventricular systolic stiffness of about 2.0 mmHg/ml of ventricular end systolic volume, while arterial stiffness is about 1.5 mmHg/ml. These low values allow relatively large changes in volume in both the heart and the vascular bed to be achieved with only modest changes in ejected pressures.
With advancing age, both ventricular and arterial stiffness increase, and these changes may be further exacerbated by common disorders such as hypertension, diabetes mellitus, and renal disease. Since the heart and arterial systems are coupled, such stiffening results in amplification of systolic and pulse pressures during ejection, faster pressure decay during diastole, and enhanced cardiac systolic loading. Stiff arteries facilitate rapid transit of the flow and pressure pulse through the vasculature, increasing the velocity at which these waves encounter regions of impedance mismatch (e.g., distal arteriolar narrowings), which then increases the amplitude of reflected pressure waves, further exacerbating systolic load. The net effect is an adverse impact on cardiac systolic and diastolic functions, with increased myocardial oxygen consumption required to provide the body with blood flow, impaired cardiovascular reserve function, labile systemic blood pressures, and diminished coronary flow reserve. Aging is also associated with endothelial dysfunction, which may in part relate to mechanical stiffening, as reduced wall distensibility can itself compromise endothelial-dependent responses to shear stress stimulation and vasorelaxation. One can consider this adverse interaction between heart and arteries as a form of coupling disease that ultimately limits the ability of the integrated cardiovascular system to respond to stress.
Abnormal ventricular-arterial stiffening and thus coupling may play an important role in patients with heart failure symptoms but with apparent preservation of systolic function. Such patients are typically older and female, with histories of chronic hypertension and a high prevalence of diabetes, obesity, and renal dysfunction. They often develop marked systolic hypertension under conditions of stress, and both their arterial and ventricular systolic pressures are very sensitive to blood volume status. While abnormal diastolic function is thought to contribute to heart failure symptoms by increasing congestion, it cannot explain the observed increases in systemic pressures, nor does it fully underlie limitations of cardiac reserve. Here, we review the pathophysiology of ventricular-arterial stiffening and its role in the syndrome of heart failure with a preserved ejection fraction (HFpEF).
PATHOPHYSIOLOGY OF LEFT VENTRICULAR-ARTERIAL COUPLING
The influence of increases in ventricular systolic and vascular stiffness on net cardiovascular function is best depicted in the pressure-volume plane. Ventricular end systolic chamber stiffness is expressed as end systolic elastance (Ees) ( Fig. 31-1 ), defined by the slope of the end systolic pressure-volume relationship. Ventricular afterload can be represented as aortic input impedance, derived from Fourier analysis of aortic pressure and flow waves. Impedance is expressed in the frequency domain and thus is more difficult to match with optimal measures of ventricular systolic function, which are typically determined in the time domain. One approach to this problem involves the development of a vascular parameter that shares units applicable to the heart, namely effective arterial elastance (Ea). Ea combines both mean and pulsatile loading, providing a lumped parameter that reflects the net impact of arterial vascular load on the heart. This index was developed and validated by Sunagawa et al. in the mid 1980s and then applied and verified in humans by Kelly et al. The latter group confirmed that the simple ratio of end systolic pressure to stroke volume (Pes/SV) could serve to estimate Ea in both hypertensive and normal humans. Graphically, Ea can be depicted as the absolute value of the slope of a line linking the coordinate points of (end systolic volume [Ves], Pes) and (end diastolic volume [Ved], P = 0) (see Fig. 31-1 ).
Coupling of heart and artery is often then depicted by the interaction of these two relations and expressed as a ratio of Ea/Ees. The intersection of these lines determines Pes and Ves (see Fig. 31-1 ). As shown in Figure 31-2 , the Ea/Ees ratio is fairly preserved with normal aging to maintain optimal efficiency, declining somewhat in women, while both the numerator (vascular load) and the denominator (ventricular stiffness) increase. Ea is dominated by mean ventricular load, namely systemic vascular resistance, but it is also altered by artery stiffening to increase pulse pressure. Blood pressure pulsatility rises with aging, and this is reflected in the pressure-volume diagram in the elderly individual (see Fig. 31-1B ) and the HFpEF patient (see Fig. 31-1C ) by the rise in systolic pressure throughout ejection (arrows) . Since Ea is determined by the ratio Pes/SV, the greater the disparity between Pes and mean arterial pressure (i.e., the more pulsatile or stiff the arterial system), the higher Ea will be relative to mean resistance load. Ea also varies directly with heart rate, since for any given cardiac SV, the systolic pressure will increase or decrease proportionally with the number of strokes (i.e., cardiac output). It is important to keep these factors in mind when interpreting data reporting on Ea and Ea/Ees coupling.
Effective coupling of heart to artery can be defined in several ways. One is the optimal transfer of blood from heart to periphery without excessive increases in pulse or systolic pressures. Another is optimal cardiovascular flow reserve without compromise to arterial pressures. One can mathematically express optimal coupling as the interaction that best enhances the work performed by the heart on the body (i.e., optimal external work). Lastly, one must consider the efficiency of the heart in performing this work—the energy consumption required to effect this external work. All of these are reasonable definitions, although prior experimental and clinical studies have tended to focus on the latter two: optimizing external work and efficiency. For this, one can both predict and observe experimentally that an Ea/Ees coupling ratio of 0.6-1.2 achieves near optimal work and efficiency. This range is normally maintained under various physiologic stresses, as shown elegantly some years back in exercising animals. It can become very high, particularly in dilated cardiomyopathy, where depressed heart function (low Ees) is coupled to a high arterial impedance (high Ea).
Ventricular-Arterial Stiffening in Heart Failure with a Preserved Ejection Fraction
In patients with HFpEF, the Ea/Ees ratio falls compared with younger individuals but is similar to that of nonsymptomatic hypertensive elderly patients. Importantly, it still falls in a range where external work and efficiency are not likely compromised. However, while the ratio itself is reduced, the absolute values of both numerator and denominator are significantly elevated ( Fig. 31-3 ). Thus, HFpEF patients have elevated vascular stiffness, as well as increased ventricular stiffness in both systole and diastole. The net interaction of ventricular and arterial stiffness is important because it can significantly affect the first two components of what we can consider optimal coupling—blood pressure homeostasis and preservation of adequate cardiovascular reserve. As displayed in Figure 31-4 , an increase in both Ea and Ees means that systolic pressures are much more sensitive to changes in cardiac preload, and thus central vascular blood volume. Small changes in volume that might accompany dietary indiscretion or diuretic usage will translate to more exaggerated changes in arterial pressure. This also predicts higher pressures during stress, which, in addition to requiring a greater amount of energy to deliver a given SV, can also alter both ejection and relaxation. A higher resting Ees means that there can be less effective contractile reserve to call upon during stress demands ( Fig. 31-5A ). Higher combined ventricular and vascular stiffening leads to greater cardiac energy demands to provide the body with a given amount of blood flow (see Fig. 31-5B ). Lastly, increased vascular stiffness and thus pulsatile loading can affect vascular homeostasis, coronary perfusion, and flow reserve. We next discuss each of these mechanisms in more detail.
Mechanisms of Ventricular-Vascular Stiffening
Ventricular stiffening is a product of both passive and active muscle properties. Passive behavior is somewhat of a misnomer, since diastolic tone is regulated in part by calcium and also by the phosphorylation state and isoforms of various sarcomeric proteins. Nonetheless, we can ascribe diastolic stiffening to properties of muscle cell size, wall geometry, intrasarcomeric protein composition, cytosolic and membrane distensibility, and extracellular matrix composition, fibrillar crosslinking, and biophysical properties. These are discussed in detail in Chapters 2 , 6 , and 30 . Systolic ventricular stiffness is related to the same determinants of stiffness in diastole, as well as activated myofilament properties, changes in structural protein behavior shortened to smaller lengths, and interactions of the activated myocytes with the matrix. As mentioned, the latter is typically measured as chamber stiffness or elastance (pressure/volume), but has also been assessed invasively by transverse indentation methods or estimated based on stress-strain analysis. Vascular stiffening also stems from structural and muscle-tone-dependent factors. Smooth muscle tone plays an important role, as does the geometry of the vessel (e.g., dilation), elastin and collagen content, crosslinking of matrix components, and other factors. This has been recently reviewed in detail.
How do these mechanisms relate to patients with HFpEF? Many HFpEF patients have left ventricular hypertrophy (LVH), concentric chamber remodeling, or both. While in some, this may develop from a primary sarcomeric protein defect (as with genetic hypertrophic cardiomyopathies), this cause is uncommon compared with the maladaptive response to chronically increased ventricular load, and we will focus on the latter here. In a large observational study of consecutive patients presenting with HFpEF, the mean LV mass index was 66.5 g/m 2 . well above cutoff partition values for defining LVH (46.7 g/m 2 . in women and 49.2 g/m 2 . in men). Myocyte hypertrophy has been documented, along with a modest rise in myofibrillar content and increases in passive myocyte stiffness in triton-skinned cells. Myocardial fibrosis is also frequently observed, although this is less clearly different from hearts displaying dilated cardiomyopathy. Ventricular cellular passive stiffness has been found to correlate with estimates of diastolic chamber stiffness in HFpEF subjects, and indeed diastolic stiffening is reported in these patients. Not all diastolic stiffening is intrinsic to the chamber, however, as external loading factors also appear to contribute importantly to observed diastolic stiffening, as indicated by invasive pressure-volume analysis.
Diastolic stiffening alone can contribute to fluid redistribution into the lungs and limit net cardiac filling, but it cannot explain why patients with HFpEF typically present with severe, uncontrolled hypertension when they develop pulmonary edema and often display marked blood pressure sensitivity to vasodilator or diuretic therapy. The latter more directly relates to increased left ventricular (LV) systolic stiffening (Ees). In addition to arterial systolic pulse pressure and stiffness, which are known to increase with age, Ees also has been shown to rise in tandem (see Fig. 31-2 ). In subjects who develop hypertension and cardiac hypertrophy, and those who further develop HFpEF, this stiffening is more pronounced over age-matched controls (see Fig. 31-1B , Fig. 31-3 ). Since Ees also has been viewed as a measure of systolic contractile function, one might conclude that this reflects enhanced contractility. However, this seems unlikely, as other parameters less dependent on chamber geometry do not increase with aging.
In addition to being older and often hypertensive, the majority of HFpEF patients are female, and this factor may also influence abnormal ventricular-arterial stiffening. Women develop concentric LVH in the setting of pressure overload more often compared with men. In addition, in a large population-based study of largely asymptomatic individuals, age-dependent increases in ventricular systolic and arterial stiffness were found to be more marked in women compared with men (see Fig. 31-2 ). This may underlie part of the increased prevalence of HFpEF in older women. The precise cause for this remains unknown but may in part relate to hormonal factors and differences in ventricular and aortic size and length. In contrast to vascular and ventricular systolic stiffening, mean peripheral vascular resistance does not show this age-dependent rise and is greater in men. Thus women are more likely to display accentuated increases in pulsatile loading with age, associated with greater coupled ventricular-arterial stiffening.
Pathophysiology of Ventricular-Arterial Stiffening
There are several important physiologic consequences of combined ventricular and arterial stiffening, as summarized in Table 31-1 , and while each of the two components confers morbidity, the combination of both leads to synergistic effects. One major repercussion is increased blood pressure lability and sensitivity to volume and vascular loading. In a normal heart-artery system, a rise in Ved results in a given rise in Pes (see Fig. 31-4A ). However, in a typical HFpEF patient, even if the coupling ratio is normal, the same increase in Ved will lead to an accentuated increase in systolic pressure (see Fig. 31-4B ). The pressure-volume area or stroke work to achieve this given SV therefore also is higher in the HFpEF patient. Conversely, a given decrease in Ved will also lead to exaggerated drops in systolic pressure in this stiffly coupled system. Similarly, an isolated increase in afterload (Ea) in an HFpEF patient will lead to a much more exaggerated increase in blood pressure because of the higher baseline Ees (see Fig. 31-4C and D ). Increased ventricular-arterial stiffening explains why systolic blood pressure is much more preload dependent in older versus younger patients (see Fig. 31-4E and F ).
