Preload is defined by Braunwald and Ross [30] as “the force acting to stretch the left ventricular muscle fibres at the end of diastole and determining the resting length of the sarcomeres”.

Returning venous blood fills the ventricle, exerting force on the heart muscle, stretching the myofibrils [30] and is one of the main determinants of cardiac performance [31–33].

The end-diastolic ventricular volume, or preload, is well reflected by the end-diastolic wall stress (preload ~ end-diastolic wall stress) [34].

Transmural LVEDP accurately reflects the effective distending pressure responsible for the length of myocardial fibres [35].

Otto Frank [36] and Ernest Starling [37] obtained a relationship between the end-diastolic fibre length and the force of contraction:

With increasing fibre length the force of contraction increases and thus the LV or RV stroke volume (SV) [36, 37] increases or, more accurately, the stroke work (SW) increases:

[38, 39]

The diastolic ventricular filling is limited by the acutely non-distensible pericardium constraining the filling ventricles and by the cytoskeleton [40–42], thus preventing the ventricles from fluid overload [43, 44] (physiological protective mechanism) as well as from pathological dilatation [41].

With an increase in resting fibre length the velocity of fibre muscle shortening increases as well [45].

Frank [36] established a linear relationship between the left ventricular end-diastolic volume (LVEDV) as a correlate of the fibre length and the force of ventricular contraction [30, 36, 37, 43].

[46]

Starling [37] reported an increase in the contraction force with increasing atrial pressures. Starling’s result is similar to that described by Frank, as long as the increase in LVEDP rep-resents a roportional increase in LVEDV (linear relationship between LVEDP and LVEDV).This is true in most healthy persons as long as the LVEDP remains within normal ranges, but in the case of high LV filling pressures and in certain pathological circumstances the rise in LVEDP is often disproportionately high in comparison to the increase in LVEDV [23, 24, 47–49].

The LVEDP may even rise without any increase in LV filling volume, producing no increase in preload, which is essential to recruit a higher SV [23, 39, 42]. Therefore, although the LVEDP rises, there may be no adequate increase in SV; in fact, there may even be a fall corresponding with the ‘descending limb’ of the Starling curve [35, 37, 39, 50]. This descending limb described by Starling is, however, an artefact of his experimental conditions.

The effective distending pressure or ‘transmural’ LVEDP is the intracavitary LVEDP (commonly just called LVEDP) minus the surrounding pressure(s) [35].

Katz, in 1965, already assumed that intracavitary and transmural end-diastolic left ventricular pressures were only equal when the pressure surrounding the left ventricular heart muscle was negligible [35]. Otherwise the external pressure must be subtracted from the intracavitary LVEDP to calculate the effective distending or transmural pressure.

[35]

Usually, the surrounding pressure has contributions of one-third by the RVEDP and two-thirds by the pericardial pressure [51, 52]:

Transmural LVEDP = intracavitary LVEDP − (2/3 pericardial pressure + 1/3 RVEDP) Under normal conditions, RAP and pericardial pressure (PP) are nearly equal [53–55] and further changes in pericardial pressure are very closely reflected by RA pressure changes [53, 56, 57].

The close relation between changes in RA pressure and pericardial pressures allows us to give a reasonable estimate of transmural pressure by subtracting RAP from pulmonary capillary wedge pressure (PCWP) [23, 53, 56]:

with CVP reflecting the ‘surrounding pressure’ [23, 53, 56, 58, 59].

There is substantial evidence that PCWP reflects LVEDP [60–62]. CVP is measured where the vena cava leads into the right atrium [58] and, as such, equals the RAP [58, 59]. Due to the very close relations between RAP and PP (r = 0.95, p < 0.005) [63] and RAP and changes in PP [53, 56, 57] respectively, CVP is a good estimate of PP [53–55, 58, 59, 63] in daily practice. Furthermore, both, CVP and RAP reflect the RVEDP [44, 59, 63]. Over a wide range, pericardial pressure, RAP and RVEDP are literally equal [64]. Tyberg [53] demonstrated that RVEDP well represents PP in ranges between 4 and 20 mmHg. However, in case of right ventricular hypertrophy when the RV is stiffened [65] and in cor pulmonale or pulmonary hypertension, RAP and RVEDP are much higher than PP. RV-failure always cause a rise in CVP [ 58 ].

In healthy persons the surrounding pressure is low (nearly zero) and an increase in preload will increase the LVEDP more than the surrounding pressure [23, 41]. Hence, the transmural LVEDP will rise along with LVEDV [23, 43, 66], increasing the preload recruitable stroke volume (work) and thus SV, as described by Frank and Starling.

In conditions where the surrounding pressure rises substantially, external constraint increases more than LVEDP [23, 24, 47, 48, 67, 68]. Transmural LVEDP and intracavitary LVEDP will differ considerably and will change in opposite directions with a fall in transmural LVEDP, lowering the preload and, consequently, the preload recruitable stroke volume (work) will decrease.

The intraventricular pressures (intracavitary LVEDP and RVEDP) are influenced by:

The LV compliance describes the diastolic properties of the heart muscle and can be depicted by the relation between LVEDP and LVEDV [51, 72] (relation between pressure and volume).

With this in mind, the discrepancies between transmural LVEDP and intracavitary LVEDP can be related, at least partly, to the ventricular compliance [73].

The ventricular compliance varies almost continuously in the critically ill, producing changes in the intracavitary LVEDP but without any corresponding change in LVEDV [69, 74, 75]. Kumar [76], however, established evidence that continuous change in the ventricular compliance is a physiological phenomenon present in healthy persons as well as in those who are unwell.

In heart failure, the compliance of the LV is almost always reduced [50], hence, increases in filling volumes cause a higher rise in LVEDP compared to a healthy heart.

The compliance is determined by factors such as muscle mass, tissue composition, elastic properties, ventricular interactions and extramyocardial conditions including pericardial structure and intra-thoracic properties [77–80].

Raised intrathoracic pressure due to pneumonia, ARDS, pulmonary oedema, etc., as well as raised intra-abdominal pressure will increase constraint, in particular on the thin-walled RV, affecting the RVEDP and PP more than the LVEDP [47].

Furthermore, the higher the LVEDP the greater the amount of external force acting on the LV, thus impeding the LV-filling, the preload, and preload recruitable SV (SW) [24, 25].

Examples of situations which alter the surrounding pressures or produce significant external pericardial constraint are:

In the case of external constraint, LVEDP markedly overestimates effective distending pressure [42].

Changes in opposite directions (transmural LVEDP ↓ and intracavitary LVEDP ↑) are explainable now, and only an increase in transmural LVEDP is consistent with an increase in LVEDV and vice versa [23, 39, 42].

Numerous publications have established that haemodynamic monitoring by PA-catheterisation measuring intracavitary (filling-) pressures fails to be an accurate guide of the preload because filling pressures do not adequately reflect the myocardial fibre length at end-diastole and, hence, the LVEDV [36, 84–86]. If the transmural pressure is used instead, then changes in the preload are accurately reflected [35].

However, the filling pressures are still one of the most important components in assessment and treatment decision-making processes in heart failure. The heart always tries to generate an adequate CO on the lowest possible LVEDP [73, 86, 87]. In heart failure patients, a therapeutic reduction of the LVEDP is correlated with improved outcome [15, 88]; hence, unloading the left ventricle and reducing the LVEDP is the therapeutic maxim that adheres to the physiology/pathophysiology of the situation [23, 24, 36, 37, 50] and improves outcome [13, 15, 24, 88]. Thus, we might do much better in our patients with severe heart failure and cardiogenic shock using the transmural LVEDP to make our therapeutic decisions.

There is, of course, a physiological optimum and maximum of fibre distension and concomitant force development (see Fig. 1.2) [89].

Furthermore, in the situation of (acute) heart failure the Frank–Starling mechanism is markedly diminished [50] and thus, in the failing heart, an increase in fibre stretch (ventricular filling) is not accompanied by the same increase in the force of contraction as in healthy persons [90].

In the failing heart the SV depends substantially on the contractility [10, 75, 91, 92] and the afterload [3, 12, 46, 93, 94].

SV is determined by venous return (responsible for the preload) and cardiac performance (con- tractility, afterload and heart rate) [31–33].

Guyton et al. [32] evaluated the relationship between total cardiac function (contractility and total peripheral resistance) and venous return:

“The actual cardiac output changes with changes in cardiac function (CF), but with changes in venous return as well”.

Indeed, as increasing ventricular filling and (thus) ventricular stretch will lead to an increase in SV, is SV basically a function of end-diastolic filling volume [95], and as such (up to an upper limit—sarcomere length of 2.2 μm) in some aspects of the amount of venous return.

Guyton plotted the relationships (total cardiac function and venous return) on one graph (see Fig. 1.3) [32].

Under most physiological conditions, changes occur simultaneously in these relationships, although one effect will be dominant [31], for example:

As such, the special interrelationship between CVP and CO has to be considered when interpreting hemodynamic conditions: Both, CVP and CO are determined by the interaction of the two functions, the cardiac function and the venous return [32, 96, 97]. RAP/CVP is not an independent determinant ascertaining CO, and “depends on CO as much as it determines it” [98]. Accordingly, CVP can be low in a person with low blood volume and normal cardiac function, but it can also be low in cases of normal volume and good cardiac function [31]. CVP can be high in cases of normal filling with impaired cardiac function, but also in cases of normal function but with fluid overload [31].

Venous return depends on mean systemic filling pressure, right atrial pressure, and vascular resistance, particularly venous tone [98]. A rise in venus tone precipitates a progression in venous return. The pressure difference between the pressure in the periphery (systemic filling pressure which is largely represented by the pressure of the extrathoracic veins [99]) and the pressure within the right atrium (central venous pressure), is usually 7–10 mmHg whereupon RA-P is normally 0 mmHg, and determines the amount of venous return [100]. Spontaneous breathing, creating negative intra-thoracic pressures, increases this pressure gradient facilitating venous return [101]. On the other hand, increases in RAP and particulary elevated RAPs oppose venous return [98].

As the level of the right atrial pressure is decisively affected by extracardiac, intrathoracic conditions, any rise in pericardial pressure (displaying pericardial constraint, as found in case of pericardial effusion or tamponade, positive pressure ventilation and pleural effusions via heart–lung interaction) will enhance RAP [102, 103]. This rise in RAP is not accompanied by, or due to, an increase in filling volume, however it will further blunt venous return [98]. Increases in ventricular afterload and myocardial ischemia affecting diastolic myocardial properties (diminishing myocardial compliance) subsequently alter ventricular filling pressures and hence RAP (LAP respectively) [95, 104]. Accordingly, no relationship exists between RAP and SV in case extracardiac, intrathoracic pressures are elevated [103, 105].

As such, the level of the right atrial pressure is determined by the pressure with which the blood distends the atrium and by the pressure effects of the pericardial, thoracic, and pulmonary adjacent structures [103, 105].

Anyway, a high CVP may be related to (a) elevated extracardiac/intrathoracic features, or may (b) indicate RV-dysfunction and/or RV outflow obstruction (namely pulmonary hypertension PH) with blood welling up in the RA, or reflects (c) both a and b. In any case, differential diagnostic considerations are implicitly required as different causes will inevitably entail different therapeutic measures [98].

While in the arterial system the pressure depends on, and is determined by, the flow and the arterial resistance (MAP = SV × SVR) [1], the venous blood flow is determined by considering volume and venous capacitance [106]:

It is the venous capacitance which dominates the venous behaviour and the central venous pressure is determined, essentially, by the venous capacity [107, 108]. It is not the venous return (as a flow), but the volume that predominantly controls basic RAP/CVP [106].

During exercise, sympathetic activity, stimulated by the reduced activity from arterial and atrial receptors, will increase venous tone and decrease venous capacitance [109]. This will increase the venous return to the heart [110] and, in case of a recruitable preload reserve (this depends on CF [10–12, 75, 91, 94]), SV will increase [36, 37]. The immediate effect of a decrease in venous capacitance is an increase in all pressures [106], including transmural RVEDP and thus RV filling, enabling the RV to increase its systolic performance [36, 37].

Fluid infusion leads to an increase in venous capacitance, lowering the central venous pressure [108, 110, 111]. A high CVP always has consequences and will limit the venous return [49].

In patients with septic shock, Stephan [112] found that, despite vasodilatation of both the arterial and venous systems [113, 114], volume loading increased the venous tone and thus the CVP significantly and to high values (>10 mmHg). This is due to a marked reduction in the compliance of the venous system secondary to stiffening of the vein walls by several sepsis-induced mechanisms [112]. Furthermore, drastic increases in CVP indicte that the ability to accommodate in case venous return has reached its limit and that blood is welling up [98].

CVP is normally 0 mmHg at rest and might increase to 2–4 mmHg during exercise [115]. The CVP is only elevated in disease states [116, 117], a CVP > 10 mmHg often reflects an elevated RV-afterload [116].

In critically ill humans [4, 69, 93, 118, 119] as well as in healthy persons [76] we know that no correlation at all exists between CVP and preload or change in CVP and change in preload. The lack of a relationship is due to the fact that, in humans, the compliance of the atria and, in particular, of the ventricles is highly variable [76]. Furthermore, preload is not the same as fluid responsiveness [120, 121], and CVP and its change poorly (do not [122]) predict fluid responsiveness [10, 75, 123, 124].

Thus, in daily practice the absolute value of the CVP and even dynamic changes in its value are very difficult to interpret and cannot be used as a valid indicator of fluid management at all [117].

In general, a CVP ≥ 10–12 mmHg has to be considered high, and most patients within this range will not respond to volume administration [44]. Bafaqueeh [125] found that 40% of patients with a CVP < 6 mmHg did not respond to further fluid administration.

Pericardial constraint accounts for 96% of the RAP, if CVP > 10 mmHg [71]: A CVP ≥ 5 mmHg [126], and particularly when exceeding 9–10 mmHg, will exert substantial constraint on (left) ventricular filling [63, 127].

Thus, an elevated CVP > 9–10 mm Hg is always pathological [116, 117], signalling that fluid administration is unlikely to be successful [125], and that diastolic ventricular interaction (DVI) [63, 127] may be present or will occur if the CVP increases further (see part 8 of this chapter).

Haemodynamic monitoring is a cornerstone in the management of critically ill patients [117]. It helps identify pathological states [13, 128] and complications of circulatory failure [13, 117] and aids restoration of normal haemodynamic parameters to prevent tissue and organ injury, to restore organ failure/dysfunction and hence to reduce mortality [117].

When faced with a compromised circulation, volume expansion is very frequently the first therapeutic measure used to improve haemodynamic status [129]. Unfortunately, only 40–70% of all patients with acute circulatory failure respond to fluid administration (SV/CO ↑) [75], which means that 30–60% of patients are not fluid responsive and volume administration may be harmful [117, 130–132]. Both, acute and chronic right heart failure [47] as well as acute left heart failure [23, 24] may deteriorate with volume loading.

Therefore a rational approach to fluid administration is needed, where the therapeutic decision is based on correctly assessed effective intravascular volume (preload) and the probable response to increased volume [117, 133]. However, the clinical tools available to evaluate the patient’s fluid status and specifically the intravascular/intraventricular filling (preload), such as jugular venous distension, crackles on auscultation, peripheral oedema, etc., are of minimal value and very poor indicators of the volume status, particularly in the critically ill patient with (cardiogenic) shock: They cannot be validated as a useful tool or basis for treatment decisions [134–138]. The only relevant clinical sign which, although still non-specific, may indicate a possible volume deficit is the heart rate. Volume deficits are usually compensated by an increase in heart rate (>90 bpm) to maintain CO in case of low SV [92, 139, 140].

In acute heart failure patients a two-minute bedside assessment [88, 141, 142] is extremely helpful to allocate the patient to one specific haemodynamic profile (wet or dry and cold or warm) with corresponding treatment regimes [88, 142–144] (see detailed information in Chap. 2). This evaluation, however, does not provide any usable information about the patient’s actual intravascular fluid status (to classify the patient as normo-, hypo-, or hypervolaemic) or whether a cold, and thus hypoperfused, patient will respond and benefit from fluids or not [28].

Hence, in addition to this useful bedside assessment, a proper assessment of the patient’s intra-vascular volume status must be carried out to clarify whether a benefit (positive fluid responsiveness) can be expected from volume expansion before fluids are given. Blind administration of intravenous fluid may be harmful through an increase in LVEDP [130], as the elevation of the LVEDP predominantly causes the patient’s symptoms to worsen [15] and, with increasing LVEDP, the patient’s prognosis [15, 16, 88].

In case of central hypovolaemia, volume administration will induce a significant increase in SV (flow) as long as a preload reserve can be recruited [130, 131, 145]. Thus, it is important to predict in a haemodynamically unstable patient whether this patient will increase his/her systemic blood flow (SV) in response to volume expansion or not [131].

Kumar [9] showed that, in healthy individuals, volume loading increases the systolic BP/LVESV ratio and the LV-SW by:

The contribution of the Frank–Starling mechanism is only mild to moderate, the contractility is the main component [9]. Kumar examined healthy volunteers and confirmed the findings of animal studies conducted in the 1960–70s [146–148]. Flow represented by SV is the original, central, and decisive parameter to be assessed when defining fluid responsiveness [75, 76, 121, 131, 145, 149].

[75, 131, 145]

In heart failure, although the LVEDV may be in the normal range, fluid administration can fail to increase the SV due to a significant reduction in contractility [10, 75, 91, 92]. Furthermore, we know that the Frank–Starling mechanism is impaired in heart failure [50, 90] and hence volume expansion may well be harmful and worsen the haemodynamic situation [117, 130–132].

Braunwald [91], and recently Michard [10, 75] have established proof that the increase in SV due to increased LVEDV depends on the contractility and pre-infusion preload (initial end-diastolic fibre length in respect to the Frank–Starling mechanism), particularly in the case of compromised cardiac function [10, 75, 91] (see Fig. 1.4).

In those patients with intermediate pre-infusion preload (normovolemia), the effect of volume loading depends exclusively on the contractility and, in the case of a compromised heart function (lower curve) in ‘intermediate preloaded’ patients, the effect of volume loading in order to increase SV, and thus CO and/or BP, is minimal and clinically not relevant [10].

Nevertheless, even in cases of cardiogenic or other types of shock, fluid administration may initially be helpful. Up to 70% of all patients in shock show a positive response (increase of blood pressure, increasing the perfusion of vital organs) when fluids are administered [150]. In non-responders we most often find that RV-dysfunction/failure with sepsis is the main underlying reason [150].

The physiological and pathophysiological facts described above demonstrate and emphasize that preload and fluid responsiveness are not the same, and this has been stressed in many published studies [117, 120, 121, 130, 151]. Therefore, as prerequisites to a positive response to fluid administration, there must exist both a recruitable contractile reserve (myocardial reserve) and an absolute or relatively hypovolaemic central vascular and cardiac system to provide a filling reserve.

An increase in SV by ≥15% due to volume administration is the most accepted benchmark confirming a positive fluid response [123, 152–154], although others define a positive response if SV increases secondary to volume expansion by ≥10% [155–158].

The force opposing myocardial fibre shortening during ventricular ejection is called afterload [30, 323–325].

Braunwald [30] states: “the load opposing LV ejection, in its simplest sense, is reflected by the systolic blood pressure”. However, the physiology is much more complex and systolic blood pressure has turned out to be a very poor reflection of afterload. Indeed, the arterial system imposes a hydraulic load on the heart, and a higher arterial load requires higher energy to eject a given amount of blood. This vascular, hydraulic load, opposing ventricular ejection is most completely described and reflected by aortic input impedance¹ (respectively pulmonary artery input impedance) [326–329].

The main parameters characterizing arterial input impedance are peripheral vascular resistance, total arterial compliance, and aortic characteristic impedance [326, 329, 331]. While peripheral vascular resistance specifies steady state conditions, the pulsatile load (pulsatile load is complex and time varying [332]) components are represented by:

The characteristic impedance outlines physical properties, such as viscoelasticity and dimensions of the large central, proximal arterial vessels (aorta, respectively pulmonary artery), and thus the contribution of elastic vascular properties to total load [327, 328, 340, 342, 343]. “Pulsatile afterload” largely includes characteristic impedance and pulse wave reflections [326, 340, 344], directly opposing ventricular ejection.

In fact, special attention has to be paid to the impact of the pulsatile elements on the total vascular load, as the intensity and the timing of reflected pressure waves change according to the elastic vascular properties (largely proximal aorta), and thus may exert a substantial impact on the vascular load the ventricle is facing [338, 345, 346]. In case of arterial vascular stiffening, as occurring with (physiological) ageing [347–349] or in hypertensive individuals [317, 350, 351], the wave velocity increases, and reflected waves return and sum up with incident forward waves, augmenting net pressure [318] and reaching the ventricle (already) in late systole, after-loading the ventricle [319, 338]. Concomitantly, aortic input impedance considerably increases [317, 319, 352, 353]. Accordingly, reflected pressure waves are shown to exhibit a substantial impact on systolic load imposed on the heart [345], and arterial stiffening is recognized to afterload the ventricle [317, 319, 354], by elevating (late) systolic load, thereby increasing systolic ventricular elastance, compromising ventricular filling, and influencing diastolic properties with raising filling pressures [347, 355]. Indeed, it has to be emphasized, that diastolic cardiac function is affected if arterial compliance decreases as in arterial stiffening [356]. Therefore, central vascular stiffening and pulse wave reflections determine late systolic arterial loading [338, 346].

The impact of pulsatile load, particular wave reflections, is even more relevant in the pulmonary circulation [341, 357]: In contrast to systemic circulation, resistance and compliance of the pulmonary vasculature are inversely releated to each other, and are evenly distributed over the complete vessel tree [358, 359]. Accordingly, an elevated pulmonary capillary wedge pressure, by decreasing pulmonary vascular resistance, enhances net RV afterload, due to increasing pulsatile load relative to the resistive one [357].

However, it is hard to obtain aortic (pulmonary artery) input impedance, as a frequency domain analysis (by Fourier method) of simultaneously measured pressures and flows is required [328, 360, 361] in order to describe the relation between arterial pressure and flow within a vessel/vessel system [327]. This is a technical challenge [360, 361] and additionally, it would be quite complicated to apply the derived frequency domain factors to daily clinical concepts and routines [330]. Sunagawa [362] made vascular properties (evaluated in the frequency domain) comparable with ventricular properties (expressed in the time domain), by lumping principal elements of vascular load (peripheral vascular resistance and total arterial compliance, characteristic impedance, and systolic and diastolic time intervals) together in (effective) arterial elastance (Ea), and as such, constitutes a close approximation of arterial load [363]. Effective arterial elastance characterizes aortic input impedance, and thus arterial load that is imposed on the ventricle [362]. This “simple measure”, which lumps together static and dynamic components of impedance, has been shown to perform well in experimental studies [362, 364]. Although dominated by the non-pulsatile load component (SVR), Ea is also altered by artery stiffening due to increased pulsatile load [365]. Ea is the most complete, and also reasonably applicable, delineation of aortic input impedance [330].

Operationally, Ea is numerically defined as the ratio of end-systolic ventricular pressure to stroke volume, and is directly related to heart rate and peripheral vascular resistance, and is inversely related to total arterial compliance (which is determined, in large parts, by the central elastic arteries) [366]:

[366, 367].

However, practically, Ea is derived from the pressure-volume relation, defined as the ratio of left ventricular end-systolic pressure (LVEDSP) to left ventricular stroke volume (LV-SV ) [362, 363]:

This equation can be further simplified: If LVESP equals systolic arterial pressure (sBP), corrected by 0.9 [363], Ea may be calculated as: Ea = LVESP/LV-SV = sBP × 0.9/LV-SV.

Normal Ea values are around 2.0 mmHg/mL [347, 363, 368, 369].

Regarding the right ventricular—pulmonary vessel system interaction, Ea-Pulm is reported to be a reliable measure of the load faced by the RV during systole, and accounts for pulmonary vascular resistance, compliance, and impedance, thus including pulsatile components of arterial load [341, 370, 371].

However, the tension the ventricular wall sarcomeres must overcome during systole in order to shorten is related to:

Accordingly, aside from the vascular properties opposing and affecting ejection, there are specific cardiac properties contributing to, and participating in, afterload characterization. As such, myocardial wall stress during contraction represents “true” afterload, because wall stress reflects both central aortic and peripheral, vascular loading conditions and intrinsic heart muscle properties, such as LV geometry, LV size and intra-cavitary pressure [324, 375–377].

The relation between afterload and systolic ventricular wall stress can be formally defined by the law of Laplace [329, 378–381]: ơ = p x r/2h

at which the ratio r/h is a main determinant of wall stress [5, 329, 374].

(ơ represents wall stress, p = ventricular pressure, r = LV or RV radius and h = wall thickness).

(The law of LaPlace applies to spherical figures, thus its transposition and application on the right ventricle with its varying regional internal radius may be problematic [380]).

Thus, directly applied:

[382]

Dilatation will induce an increase in LV(RV) diameter and generally in LV (RV) filling pressure, and as such leads to a rise in wall stress. An increase in wall thickness (in the case of hypertrophy) reduces the wall stress.

[382, 383].

Determinants of the LV wall stress mediated by LaPlace’s law are continuously changing during systole, producing varying measurements of LV wall stress depending on the phase of the cardiac cycle. Peak wall stress occurs within the first third of ejection, and wall stress then declines to its end-systolic value, which is less than 50% of the peak value. At the same time, the total systolic wall stress (estimated by the stress time integral), predicts myocardial oxygen consumption [384].

All measures show a significant difference and the choice of index depends on the question being asked [384]:

The very good correlation between end-systolic wall stress and myocardial fibre length at end-systole [69], as well as between end-systolic wall stress and end-systolic ventricular volume (ESV) [385–387], underlines the fact that the end-systolic wall stress is literally the (after)load that limits the ejection [388, 389].

[382, 385–387].

Furthermore, several authors have confirmed the excellent correlation between end-systolic wall stress and LV afterload in daily practice [323, 376, 386, 387, 390].

During systole, the LV-chamber size will decrease while the ventricle contracts, and thus the wall tension will fall. When the afterload increases, a greater rise in pressure is necessary for any given reduction in chamber size, and therefore, wall tension during systole is higher. The pressure increase has to be even greater, of course, in a primarily dilated LV [19].

By the way, there are two echocardiographic methods described by Reichek [378] (M-mode assessment, meridional wall stress) and Greim [391] (2D-assessment, circumferential wall stress), which directly assess the end-systolic wall stress. Both are time consuming, require advanced skills, and Greim [391] expresses concerns about the ability of the M-mode method to recognize acute changes in afterload in patients during cardiothoracic surgery.

To summarize, two alternative biophysical concepts may describe and characterize afterload [329]. Ross and co-workers [392] gave evidence, that the level of wall stress, rather than input resistance or pulsatile impedance, determine ventricular performance, favouring wall stress as the most exact feature representing “true” afterload. Furthermore, arterial input impedance specifically refers to vascular properties, while more or less neglecting cardiac properties. As such, while wall stress integrates the forces that oppose ventricular ejection, Ea is a measure of the hydraulic load faced by the ventricle [380, 393]. Wall stress is considered to be the most accurate feature to describe ventricular afterload [329]. However, while wall stress estimation has not gained any clinically feasible relevance in daily practice, Ea may be obtained with reasonable effort.

In order to perform well, heart performance and afterload have to match, leading to the concept of afterload mismatch [329]: Basically, “a mismatch can be induced acutely in a normal heart if end-diastolic volume is not allowed to compensate for the increase in afterload” [394]. Subsequently, SV, EF and ventricular circumference will fall [329]. Examples include: volume depletion in the presence of rapid and substantial rise in systemic pressure, and increases in afterload in a ventricle already having utilized the maximal preload reserve with average sarcomere length exceeding 2.2 μm, indicating maximal stretch (limited by pericardial constraint, and as such explaining why there is no descending limb of the Starling curve). Imposing an extra load on such a ventricle will cause a sharp drop in SV, unless the contractility can be increased intrinsically or by applying inotropic agents [395].

Afterload can be defined as the forces that oppose ventricular muscle fibre shortening [329, 396]. Features opposing comprise:

Systolic wall stress integrates the forces opposing ventricular ejection, accordingly, wall stress may be considered the most accurately feature characterizing the load faced by the ventricle during systole [324, 332, 375–377, 397].

However, the clinical feasibility to apply the one, or the other, method assessing the highly complex relationship between ventricular contraction, arterial system and blood flow at bedside is nearly impossible [329]. Furthermore, that one single parameter can fully encompass all aspects cannot be expected [329]. To simultaneously measure pressures, wall thickness and radius in a constantly changing system, in order to determine wall stress, is currently not feasible. Therefore, simplified derivations are necessary and as such, effective arterial elastance has gained high acceptance and is widely used, at least in medical research, to evaluate the discussed relationships, and thus get insights into pathophysiological and pathogenetic processes, interrelations, and sequences [329].

In daily clinical practice, the systemic (peripheral) vascular resistance (SVR) is the most common parameter used to describe the actual afterload, and often, SVR is used synonymously with after- load.

SVR however, only reflects the non-pulsatile component of the peripheral load under steady state conditions [398]. It does not comprise the impact of wave reflections, arterial impedance, or ventricular properties. Each of these phenomena affect LV-afterload independently of peripheral vascular resistance or arterial pressure [376]. Ageing, hypertension, and aortic stiffening contribute considerably to the pulsatile component of the afterload, and thus, this component becomes more prominent under those conditions [399, 400]. Lang [401] showed in his investigation that the measurement of SVR substantially underestimates the change in afterload when LV afterload alone was decreased, increased, or remained unchanged, but with a simultaneous increase in contractility. These findings are not surprising, because from the peripheral pressure-flow relationship, the systemic peripheral resistance is not seen by the LV [378]. Nevertheless, SVR accounts for 90% of the resistance to ejection (arterial resistance is the dominant component of impedance load [402]) [403] and thus is justified as being the most commonly used parameter to clinically estimate afterload [222]. Furthermore, SVR may be very helpful in clarifying the diagnosis [13, 128, 222], particularly in hypotensive patients, and in heart failure syndromes, as shown by Cotter [13].

The fundamental pathophysiological alteration in acute heart failure syndromes is a substantially and inappropriately elevated afterload, with a markedly elevated systemic resistance/markedly increased LV outflow impedance, exerting a high (end-)systolic load on the LV during ventricular ejection [11, 19, 394]. This is referred to as afterload mismatch, defined by “a fall in SV due to inappropriately high afterload” [329, 394, 404]. In heart failure syndromes, the LV afterload becomes the decisive determinant of cardiac performance [11, 12, 14]. As early as 1977, Cohn and Franciosa published their impressive diagram showing the correlation between afterload and cardiac performance/cardiac output (SV) (see Fig. 1.6) [11].

SV depends decisively on the magnitude of the afterload [3, 46, 93]. Furthermore, an elevated (after)load causes an increase in the LV filling pressure [405], and thus affects the already compromised diastolic properties of the heart, resulting in a further reduction of the LV filling rate [406, 407]. Afterload is inversely proportional to the stroke volume, SV ~1/afterload [394], and therefore an increase in afterload should result in a fall in SV and ejection fraction (EF) [378, 408]. However, in healthy hearts, despite an increase in wall tension due to the increased afterload, normal fibre shortening is accomplished by a compensatory increase in contractility [9, 93]. In the case of impaired LV function the increase in afterload is not tolerated, fibre length shortening is impaired, and a decrease in EF results [19].

Finally, remember the following [382, 383]:

Contractility is defined as the inherent capacity of the myocardium to contract independently of changes in pre- and afterload [413]. This capacity of Intrinsic Force of contraction is called Contractility or Inotropy [414, 415].

Braunwald writes, “Changes in cardiac performance independent of alterations in pre- and afterload are caused by ‘contractility’. It has to be separated from changes in the performance due to a change in loading conditions” [413].

The sympathetic tone plays an important role in the regulation of contractility. The positive inotropic effect of increased sympathetic tone enables the heart, without a change in diastolic filling (without a change in the preload), to eject a higher SV or to maintain SV in case of increased afterload or increased resistance to ejection [416]. Kumar [9] found in healthy volunteers that the increase in SV due to volume loading is predominantly caused by an increase in contractility and only in minor part by the Frank-Starling mechanism, hence confirming previous results [146, 147]. Due to the increase in ‘intrinsic’ contractility, the end-systolic volume will decrease [9].

It is very difficult to measure and to express contractility as a single, independent parameter. At the sarcomere level, contractility and load are interrelated; thus, they are not independent variables [417, 418]. Any parameter attempting to characterise ‘true’ contractility has to be independent of changes in pre- and afterload, LV-size and geometry and LV-pressure [419].

The rate of LV intraventricular pressure rise dp/dt, an index of the isovolumetric phase of the contraction [420], correlates well with the LV contractility [421]. The highest dp/dt, called dp/dtmax, throughout systole is expected to be proportional to the contractility [421]. Dp/dtmax, is sensitive of preload, but not of afterload because it is measured before the aortic valve opens [421].

Dp/dtmax shows reasonably good sensitivity to detect and express changes in the ‘true’ inotropic status (intrinsic contractility) [422, 423]. It is the most valuable parameter to measure and express inotropy [422–425] and is currently the gold standard in representing the ‘true’ (intrinsic) contractility [426].

The contractile conditions of the ventricle are influenced by intrinsic properties of the ventricle at end-systole, the chamber elastance (Ees). These contractile properties of the ventricle can be quantified by the relationship between end-systolic left ventricular pressure (LVESP) and the end-systolic left ventricular volume (LVESV) [388, 427].

The ventricular pressure-volume relationship at end-systole is linear (at least under physiological conditions [428]) and its slope, Ees, quantifies the ventricular (systolic) contractile properties [388, 428, 429] (read more in part 1.9.3 of this chapter) (Fig. 1.7).

E_es is defined as LVESP divided by LVESV, thus

[430]

E_es is roughly load-independent [388], and Kass [431] found that over a wide range of load, E_es is a powerful index of true LV-contractility [432–435].

When describing the systolic properties of the heart, we must differentiate between indices referring to the ‘true’ contractility and to other parameters describing the systolic function of the heart muscle or the heart performance. The latter two are less independent than the other indices and haracterise the heart function in a more ‘global’ way. (For an overview see Baicu [368]).

LV systolic performance is characterised by the stroke work, taking into account that the heart has to generate pressure and flow (SV) [4, 93]:

[39, 436]

Normal values: 58–104 gm⁻¹ m² [437]

The systolic performance is influenced by load and ventricular configuration [438]; thus, it is not the same as contractility. Hence, abnormal performance may be present although contractility is normal (i.e. in case of high afterload) and vice versa, performance may be normal although the contractility is impaired (i.e. sepsis, MR) [438].

Whilst cardiac work describes the transferral of energy from the cardiac contraction to the de- velopment of blood flow [128], cardiac power output (CPO) describes the amount of energy generated by the heart that the whole systemic vasculature receives at the level of the aortic root [128]. Thus, it characterises the recruitable reserve still available in case of acute failure or shock in order to maintain the perfusion of the vital organs and hence reflects the severity of the patient’s illness [132]. CPO has shown substantial prognostic power [128, 439] across the broad spectrum of acute heart failure syndromes and, in particular, in cardiogenic shock [128]. CPO is defined [128] as

and follows the physical rules of fluids. Reflecting the essential task of the heart (to generate pressure and flow) [3, 4] CPO is a measure of cardiac pumping by coupling both pressure and flow domains [440].

Furthermore, CPO and its index, CPI, have shown superiority in determining the exact diagnosis of the actual heart failure syndrome compared to CI, BP, PCWP and their combination [13, 128]. Whilst the traditional haemodynamic measures and their presumed target values used in treatment protocols have been misleading [441], they have also failed to show any relevant effect when therapy was titrated upon reaching these values [442].

CPO appears to be a better parameter than CPI for predicting outcome. Adjustment of CPO for body size, yielding CPI, showed a weaker association with mortality [443, 444]. A CPO ≤ 0.53 most accurately predicts a high likelihood of in-hospital mortality [128, 439].

Conventionally SVI and SWI were used as powerful predictors of short term mortality in cardiogenic shock complicating AMI [445], but the use of CPO is now thought preferable.

The LV systolic function of the heart can be described in a number of ways but, ejection fraction (EF, %) is still the most frequently used parameter. EF is determined by the interaction of arterial and ventricular properties and is dependent on the afterload, and thus it is not exclusively governed by the LV [347, 368, 446].

[447, 448].

However,

[378, 408].

As such, EF may, by all means, be considered as a good coupling parameter, describing fundamental aspects of ventriculo-arterial coupling [449, 450] rather than contractility.

EF is thus far from being an ideal parameter to assess contractility. EF depends on afterload as well as on preload and heart volume or mass [394, 423, 451, 452].

EF will fail to report:

Normal values EF > 55% [447, 448, 457–461]; an EF > 40% is considered reasonable [457–462].

Despite its shortcomings, Braunwald [438] and Gillebert [463] state that EF is the best parameter to describe overall contractility in comparison to all others currently in use.

‘True’ LV-contractility is best expressed by:

E_es <1 mmHg/mL is found in dilated and failing hearts [464], in case of hypertrophy there will be a significant increase—up to 4 mmHg/mL [465].

It has to be stressed that CI is not an index of contractility, but rather a measure of cardiovascular flow: CI is affected by contractility, vascular stiffness and resistance, intravascular volume and filling pressures [128]. Furthermore, there is no normal CO/CI, since metabolic demands can vary widely [117].

Medications able to increase the myocardial contractility are called inotropes. In recent years the administration of inotropic drugs has been overshadowed by clear and growing evidence of adverse events and increased mortality [143, 466–470], particularly when given in patients with reasonably preserved left ventricular function (EF > 40%) [471, 472]. Conners [473] and Sandham [474] found a significantly increased mortality when clinically stable patients were treated with conventional inotropic agents secondary to numerically low cardiac output. Only patients who absolutely require inotropic support secondary to low output as result of severely impaired contractility and who are resistant to other treatments should be treated by such drugs [462, 475].

The European Society of Cardiology (ESC) recommends inotropic agents in heart failure syndromes if the illness has deteriorated to become life-threatening and the situation has become critically dependent on the haemodynamics: “Inotropic agents are indicated in the presence of peripheral hypoperfusion with or without congestion or pulmonary oedema refractory to diuretics and vasodilators at optimal dosages” [462].

The right and the left ventricle are anatomically and functionally closely interrelated, since they share the interventricular septum (IVS), the pericardium, and (by their continuity) parts of myocardial fibres [482]. Ventricular interdependence characterizes the “response of one ventricle to the changes in pressure and volume of the other” state Elzinga et al. [483]. However, the interactions include even more features as they refer to the changes in size (volume), shape, pressure, and (concomitantly) compliance of one ventricle due to direct, predominant mechanical (independent of neurohormonal and circulatory effects ), impact on the other [482], and further to a systolic contribution of the LV to RV contractile performance [393, 484]. Accordingly, diastolic ventricular interaction largely refers to the competition of the two ventricles for space within the non-distensible pericardial sack, namely when RV dilates, and systolic ventricular interaction applies to the contribution of LV to RV systolic performance (read more about this issue in Chap. 4) [393]. The impact may be even dramatic in case of acute changes in RV size and pressure [24, 41, 47, 485].

These diastolic interactions are mediated via the shared structures of the two ventricles, the interventricular septum and the pericardium. The pericardium has constraining effects on ventricular filling due to its poor distensibility and its pressure transmitting effects [47, 486]. The interaction mediated by the septum and the pericardium is called ‘direct’ interaction, compared to the so called ‘series’ interaction which refers simply to the physical relation between the two ventricles and their outputs: The two ventricles are coupled in a row, one after the other and thus their output necessarily has to be equal over time [31, 41].

The shape of the septum, under physiological conditions, is concave when viewed from the LV side. There is no difference during systole and diastole, due to the fact that the LVEDP always remains higher than the RVEDP and increases proportionately during systole [41]. Kingma established proof that the position of the septum is determined by the end-diastolic pressure gradient between LV and RV [487]:

[487].

In disease, the position of the septum can change markedly due to changes in the pressure gradient, which will alter the end-diastolic volumes substantially [47, 48, 119, 487, 488]. In acute RV pressure or volume overload Kingma showed that the interventricular septum becomes flattened or even convex at end-diastole due to RV dilatation and raised RVEDP, diminishing the transseptal pressure gradient and pushing the septum towards the left ventricle [487]. Numerous publications confirm the change in the septum position in different diseases such as acute and chronic pulmonary hypertension [47, 48, 489], congestive heart failure [23, 24], and mechanical ventilation [119].

This leftward shift of the septum contributes significantly to the reduction in LV-filling; thus, total LV-volume and end-diastolic volume are reduced and the SV will fall as a consequence.

The very poorly distensible pericardium supports this process by exerting constraint, restricting the total heart volume from changing [478, 490].

All cardiac chambers (except the posterior part of the LA where the pulmonary veins enter) are enclosed by the pericardium. It works as a tight, unyielding band around the minor axis of the heart, fixing the cross sectional area of the heart and causing direct ventricular interaction [491].

Thus, an increase in the cross-sectional area of one ventricle, e.g. due to volume loading or enlargement, necessarily reduces the area of the opposite ventricle with less filling potential, causing an increase in the pericardial pressure, and altering the transmural pressure [23, 491]. The total cardiac volume remains unchanged [488, 490].

Increasing pressures in the pericardial space will exert a progressive restraining effect on ventricular filling, termed pericardial constraint [488]. When the pericardium becomes stretched due to enlargement of the ventricles, such as in chronic heart failure or due to volume loading, the filling—in particular the left ventricular filling—becomes significantly restrained [23, 127]. With further stretch the pericardium is even less distensible [486] and, especially in cases of acute change, the pericardium, with its constraining effect, plays a key role in loading conditions [85, 492–494]. Under those conditions the pericardial pressure (PP) will increase progressively and will significantly constrain the filling. PP rises in an exponential manner [491] and once the pericardium becomes ‘overstretched’, an exponential increase in LVEDP is seen [83, 495].

Raised intra-thoracic pressure, e.g. due to raised intra-abdominal pressure, chest infection, etc., will affect, secondary to an increased constraint on the thin walled RV, the RVEDP more than the LVEDP (rise in RVEDP > rise in LVEDP) [47, 48]. Hence, the transmural LVEDP (= LVEDP − RAP/CVP) will decrease with less LVEDV and less LV end-diastolic fibre stretch, and a reduced SV will result.

Ventricular interaction due to pericardial constraint is diminished as long as the PP is <5 mmHg [126]; when exceeding 9–10 mmHg the pericardium will exert a significant constraint on ventricular filling [63, 127]. Furthermore, when intraventricular LVEDP exceeds 10(12)–15 mmHg, the LVEDP-LVEDV relation becomes much steeper and the pericardium limits further increases in LV volume [83, 133, 495].

In pulmonary hypertension fluid administration is shown to increase RVEDP more than LVEDP [47, 48]. The concomitant (along with RVEDP) increase in pericardial pressure will exceed the rise of the LVEDP (due to a higher increase of RVEDP compared with LVEDP), thus transmural LVEDP and therefore LV-preload will be reduced due to pericardial constraint [23, 51] (Fig. 1.8).

Fluid administration in pulmonary hypertension