Pathophysiology of the right ventricle in health and disease: an update

Abstract

The contribution of the right ventricle (RV) to cardiac output is negligible in normal resting conditions when pressures in the pulmonary circulation are low. However, the RV becomes relevant in healthy subjects during exercise and definitely so in patients with increased pulmonary artery pressures both at rest and during exercise. The adaptation of RV function to loading rests basically on an increased contractility. This is assessed by RV end-systolic elastance (Ees) to match afterload assessed by arterial elastance (Ea). The system has reserve as the Ees/Ea ratio or its imaging surrogate ejection fraction has to decrease by more than half, before the RV undergoes an increase in dimensions with eventual increase in filling pressures and systemic congestion. RV-arterial uncoupling is accompanied by an increase in diastolic elastance. Measurements of RV systolic function but also of diastolic function predict outcome in any cause pulmonary hypertension and heart failure with or without preserved left ventricular ejection fraction. Pathobiological changes in the overloaded RV include a combination of myocardial fibre hypertrophy, fibrosis and capillary rarefaction, a titin phosphorylation-related displacement of myofibril tension–length relationships to higher pressures, a metabolic shift from mitochondrial free fatty acid oxidation to cytoplasmic glycolysis, toxic lipid accumulation, and activation of apoptotic and inflammatory signalling pathways. Treatment of RV failure rests on the relief of excessive loading.

1. Introduction: Relevance of the right ventricle

The right ventricle (RV) has long been thought to be dispensable. In 1943, Starr et al. showed that ablation of the RV in dogs is compatible with life with little change in venous pressures.¹ In 1971, Fontan and Baudet introduced the cavo-pulmonary anastomosis as a palliative intervention for certain cardiac malformations.² So-called Fontan patients indeed enjoy a near-normal sedentary life without RV for several decades.³ However, they have a 30–50% decrease in aerobic exercise capacity.⁴ This indirectly discloses the contribution of the RV to maximum cardiac output.⁴ Furthermore, any cause increase in pulmonary artery pressure (PAP) in Fontan patients is a cause of circulatory failure.³ Thus, the RV may be dispensable but only in a quiet sedentary life, with a normal cardiac output and a low-resistance pulmonary circulation.

The RV in mammals and in birds is a thin-walled crescent shape volume generator coupled to systemic venous return on one side and to the pulmonary circulation on the other side. This structural particularity is the result from a Darwinian selection process.⁵ Evolution from poikilothermic reptiles and amphibians to endothermic mammals and birds requires an increased gas exchange to meet metabolic needs. Sufficient oxygen and carbon dioxide exchange are only possible through an extremely thin blood–gas barrier. Such a structure is efficient but fragile. It is therefore protected from high systemic vascular pressures within a separate low resistance pulmonary circulation.⁵

The structure of the RV is not designed to cope with brisk increases in PAP. In anaesthetized animals, pulmonary ensnarement to a systolic pressure exceeding 60 mmHg is associated with extreme RV dilatation and circulatory collapse.⁶ However, if given time, the RV adapts by an increased contractility. This ‘homeometric' or inotropic adaptation is turned-on within 3–5 min and is not therefore dependent on hypertrophic remodelling. In Starling's heart–lung preparation, an acute increase of either preload or afterload is immediately associated with an increase in end-diastolic volume and end-systolic volume (EDV and ESV), a decrease ejection fraction (EF), and a preserved stroke volume (SV).⁷ After several minutes, this ‘heterometric adaptation', which conforms to so-called Starling's law of the heart, is replaced by a homeometric adaptation as EDV returns to normal, ESV is decreased, EF is increased, and SV is back to baseline.⁸ It is now better understood that Starling's law of the heart applies to beat-by-beat adaptations, for example with changes in body position or when the homeometric adaptation eventually fails in patients with chronically overloaded hearts.

In spite of embryologic and structural differences, homeometric and heterometric adaptations to changes in loading conditions are basically the same for the RV and for the left ventricle (LV).⁹ However, contractility responses to increased preload or afterload vary in magnitude depending on time course of increased loading, volume status, reactive ventricular hypertrophy, the presence or not of systemic diseases, and yet unknown differences in molecular signalling pathways. The progression from homeometric adaptation to predominant heterometric adaptation is not an on–off phenomenon but evolves in variable proportions. In permanently increased loading conditions such as in severe pulmonary hypertension (PH), functional state, exercise capacity, and cardiac output may remain reasonably well preserved at rest, while right heart dimensions progressively increase along with a decreased EF, altered exercise capacity, and shortened life expectancy.^10,11

The RV has attracted a lot of attention in recent years with accumulation of evidence of its prognostic relevance in severe PH, any cause heart failure (HF), and systemic diseases. The present review aims at an update of gold standard vs. clinically applicable methods to assess RV function with consideration of new pathophysiological insights and a selection of basic science advances of translational pertinence.

2. RV overload

Ventricular function is determined by preload, afterload, and contractility. While RV preload is easily defined as EDV, which is the moment in the cardiac cycle when myocardial fibres are maximally stretched before the isovolumic contraction, there are several equally valid but conceptually different definitions of RV afterload.^10–12

The first is maximum wall tension, or the maximum value of the product of ventricular volume by ventricular pressure, divided by wall thickness (T) as an application of Laplace's law for spherical structures.

Maximum wall tension is increased by volume loading as well as by pressure loading. Starling taught that preload and afterload are as inseparable as the ebb and the flow. However, while maximum wall tension may be considered as the reference definition of afterload, it is difficult to apply to the RV because of its irregular shape and inhomogeneous contraction.

The second is the assessment of the forces that oppose RV ejection by hydraulic power (W_TOT), calculated from the integration of instantaneous pulmonary arterial (PA) pressure and flow waves. Total power W_TOT is the sum of steady power (W_ST), which defines the work of mean forward flow, and oscillatory power (W_OSC), which defines the additional work entailed to oscillations.

W_ST is equal to the product of mean PAP (mPAP) by SV.

Because of the near-constancy of PA compliance to pulmonary vascular resistance (PVR) relationship, or ‘time constant' of the pulmonary circulation, W_OSC can be calculated to be stable at 23% of W_TOT,¹³ and therefore W_TOT is estimated at 1.3 times mean power W_ST.

However, this estimation carries uncertainty because of inevitable errors on mPAP calculated from systolic (sPAP) and diastolic or derived from the integration of a PAP wave, and imprecision of the time constant calculation.^13–15

The third is pulmonary vascular impedance (PVZ) or oscillatory PA pressure as a function of oscillatory PA flow.¹⁵ A PVZ calculation integrates all the forces that oppose PA forward flow. The determination of PVZ requires high-fidelity instantaneous pressure and flow measurements and a frequency analysis. The result is graphically represented by a PVZ spectrum or pressure-flow ratio and phase lag as functions of frequency. Relevant numbers of the spectrum are 0 Hz impedance (Zo), or mPAP divided by cardiac output also called total PVR, high-frequency impedance also called characteristic impedance (Zc), first minimum and maximum of the pressure-flow moduli and low-frequency phase. In this analysis, Zo is sensitive to changes in peripheral resistance vessels, while Zc is sensitive to proximal PA elastance. Measuring a PVZ spectrum is unpractical at the bedside because of the needs of expensive and sophisticated catheter technology and results expressed as several numbers in the frequency domain most clinicians are not familiar with. None of the numbers defining a PVZ spectrum can be considered as a sole measure of afterload.

The fourth is arterial elastance (Ea) or RV end-systolic pressure (ESP) divided by SV.

Ea measures afterload at the moment of the cardiac cycle when RV elastance is maximal.^10–12 Maximum elastance as a gold standard estimate of contractility in an intact animal occurs before end-systole in the normal RV but coincides with end-systole in the RV coupled to a hypertensive pulmonary circulation. End-systolic elastance (Ees) is generally an acceptable approximation for maximum elastance of the RV.^10–12

It is sometimes assumed that mPAP could be a reasonable approximation of ESP, so that,

or total PVR divided by heart rate.

However, in patients with PH, mPAP under-estimates ESP in proportion to increased systolic pressures roughly following the relationship.¹⁶

Because of relatively wide scattering of data points and insufficient data from patients with normal or mildly hypertensive pulmonary circulation, it is preferable to calculate Ea from directly measured ESP. Defining ESP requires the determination of a PV loop.

3. Coupling the RV to the pulmonary circulation

Measurements of either right or LV Ees and Ea can be obtained from a family of pressure–volume loops at several levels of preload^17,18 or from single beat ventricular pressure and flow output measurements.^19,20 This is illustrated in Figure 1. The single beat method is convenient as it avoids technically demanding measurements at variable levels of loading and may be preferred for the RV to circumvent the difficulty of absolute ventricular volume measurements.

Single beat RV-PA coupling relies on a maximum RV pressure (Pmax) calculation from the non-linear extrapolation of isovolumic portions of a ventricular RV pressure curve (thus corresponding to peak RV pressure during a non-ejecting beat) and measurement of P as a function of V during RV ejection.²⁰ A straight line tangent to upper left shoulders of a family of PV loops or from Pmax tangent to a single PV curve defines Ees. A straight line drawn from the Ees to end-diastolic elastance (Eed) points of any single PV loop defines Ea (Figure 1).

The coupling of RV function to the pulmonary circulation in PA hypertension (PAH) was first reported with single-beat calculation of the Ees/Ea ratio from magnetic resonance imaging (MRI) measurements of RV volumes and standard fluid-filled catheter measurements of RV pressures in six stable patients.²² The striking finding of that study was that Ees was increased, but insufficiently so to match an even more important increase in Ea, so that Ees/Ea was halved as compared to controls indicating insufficient contractility (‘homeometric') adaptation to afterload. Yet, EDV and ESV were not increased, indicating adequate compensation of RV function in keeping with considerable Ees/Ea reserve. These results have been confirmed in small cohorts of patients with either idiopathic or systemic sclerosis (SSc)-associated PAH or chronic thromboembolic PH (CTEPH) and in one patient with a systemic RV.^23–28 In these studies, Ees/Ea was measured using gold standard micromanometre-tipped catheter technology and the single beat method^{22,23,25,26,28} or the multiple-beat method using a decrease in venous return induced by a Valsalva manoeuvre.^24,27 Exercise decreased Ees/Ea in PH patients, in contrast to preserved Ees/Ea during exercise in controls.²⁶ Exercise-induced decrease in Ees/Ea was associated with an increase in EDV, which was typically observed in SSc-associated PAH.²⁷ These clinical studies confirmed observations in experimental animal models of HF, sepsis, and PH.¹⁰

How does the overloaded RV evolve from adaptation (preserved volumes) to failure (increased volumes)? The critical levels of decoupling associated with onset of increased dimension (‘heterometric') adaptation and eventual clinically identifiable congestion in patients with severe PH are not exactly known.^10–12 The optimal Ees/Ea ratio for ejection at minimal energy cost was initially modelled to be higher than 1, actually in the range of 1.5–2.¹⁰ However, the limits of normal of the Ees/Ea ratio have not been exactly defined in clinical studies. In experimental animal models of embolic PH, the RV dilated when the Ees/Ea ratio decreased to 0.7–1.0.^29,30 It was recently shown in a study combining experimental animal observations and measurements in patients with CTEPH that the Ees/Ea ratio could be decreased down to 0.68 before the onset of decreased SV.²⁸ These observations have been confirmed.²¹

Thus, while being gold standard measures of contractility and afterload, Ees and Ea and Ees/Ea determinations may not be sensitive to clinical deterioration until advanced stages of PH.^10,11 On the other hand, the persistence of increased Ees and maintained Ees/Ea in severe PH makes the measurement insensitive to efficient therapeutic interventions. Epoprostenol in experimental PH decreased Ees along with decreased PVR, but this did not indicate a negative inotropic effect as the Ees/Ea ratio was maintained.³¹ A similar observation was reported in PAH patients on chronic prostanoid therapy, in whom PVR and RV volumes decreased, SV and EF increased, and indices of diastolic function improved in spite of decreased Ees and unchanged Ees/Ea.³² These observations may be explained by the active adaptation of contractility to afterload, by yet unknown mechanisms.³¹ An alternative explanation is in the non-linearity of end-systolic PV curves.³² These have been shown to present with increased slopes when ventricular pressure increases, so that Ees decreases along with the unstressed ventricular volume with effective PH therapies.³³

The PV loop has of the LV a square shape with easily identifiable Ees at its upper left corner.¹⁷ The normal RV PV loop has a rounded triangular shape with early systolic peaking of pressure and persisting ejection after Ees which is therefore more difficult to discern.¹⁸ However, the shape of RV PV loops changes with progression of PH, from a triangular shape with early systolic peaking of pressure, to trapezoid and triangular shapes with late systolic peaking of pressure and mid systolic dip of pressure (‘notching').^34,35 These morphological changes have been shown to be of prognostic relevance.³⁵ The coupling of RV function to the pulmonary circulation assessed by Ees/Ea is better preserved at similar severity of PH in female compared to male patients.³⁶ In general, the Ees/Ea ratio is an independent predictor of outcome in PH patients, with a cut-off value for prognostic discrimination of 0.65 to 0.8.^37,38

Single beat and multi-beat measurements of Ees and Ea produce essentially the same results.³⁷ However, because of sympathetically mediated increase in Ea at reduced cardiac output²⁰ and the non-linearity of end-systolic P-V relationships,³³ this may not be confirmed at lower or higher than normal cardiac output.

High-fidelity measurements of instantaneous RV pressures and volumes and derived elastance calculations have contributed to current understanding of RV function adaptation to loading conditions in a variety of cardiopulmonary conditions but are technically demanding and therefore not suitable for daily bedside clinical practice.³⁹

4. Simplified measurements of RV-arterial coupling at the bedside

Since the Ees/Ea ratio has pressure as a common term, it can be simplified to a ratio of volumes, easy to measure with MRI⁴⁰:

Simple modelling predicts that ESV and EDV increase above the upper limits of normal in a SV-dependent fashion when SV/ESV decreases below 54% and SV is normal.⁴¹ Since SV is equal to EDV−ESV and thus SV/ESV = EF/(1−EF), this corresponds to a RV EF of 35%.⁴¹ It has to be noted that because of their curvilinear relationship, SV/ESV and EF are more sensitive to early and late stages of RV dysfunction, respectively.⁴¹

It is interesting that both SV/ESV and EF of 54 and 35% are effectively associated with RV volumes above the upper limit of normal²¹ and correspond to cut-off values discriminating poor from favourable outcome in severe PH.^42,43

The Ees/Ea ratio can also be simplified to a ratio of pressures^42,44:

Based on the above-developed equations, RVEF can be derived from a RV pressure curve⁴⁵:

Pressure-derived EF has been validated against MRI-derived EF, with Bland and Altman analysis showing no or minimal bias (a measure of accuracy) but relatively large limits of agreement (a measure of precision).⁴⁶ Both RV volume- or pressure-derived estimations of Ees/Ea have been shown to provide reasonably accurate but with insufficient precision for individual assessments.³⁷

Simplified measurements of the Ees/Ea ratio rest on incorrect assumption that the end-systolic PV relationship is a straight line crossing the origin.⁴⁴ Thus, single point SV/ESV or EF may be dependent on the absolute level of P or V. Furthermore, assuming zero-crossing of PV neglects the unstressed volume of the RV, which may be of clinical relevance.³² Replacement of ESP by easier to measure mPAP in the pressure-derived equation has been proposed in some studies.^26,44 However, as already mentioned, mPAP under-estimates PES in proportion to PH severity.^16,35

5. The importance of volume measurements

The assessment of RV-PA coupling is easier in clinical practice using non-invasive imaging modalities. The measurements are preferably focused on RVEF because of its relevance to Ees/Ea and conceptual familiarity with EF measurements in left HF. Several studies have demonstrated that MRI-determined RVEF is an independent predictor of outcome in severe PH, with a cut-off value of 35–40% down from normal values around of 60–70%.^47–50

An interesting alternative to cardiac magnetic resonance (CMR) is three-dimensional (3D) echocardiography. While RV volumes measured by 3D echocardiography tend to be slightly lower than those measured by MRI, derived RVEFs are within the same range and have been shown to be of prognostic relevance in patients with HF⁵¹ or with PAH.⁵² As 3D echocardiography is not yet widely available, two-dimensional (2D) echocardiography remains a useful alternative with reasonable validation of fractional area change (FAC) by high correlation coefficients against RVEF.⁵³ The FAC has been shown to be predictive of outcome in PH, also with a rigorously determined cut-off value of ∼35%.⁵⁴

Reversing increased RV dimensions or ‘reverse remodelling' by targeted therapies may restore PAH life expectancy to normal.^55–58 Reverse RV remodelling with improved EF or FAC in severe PH is related to decreased PVR in a sigmoid fashion, with the steepest portion of the curve at a % decrease in PVR around 50%.⁵⁵ Such a decrease in PVR in PAH requires the combination of at least two therapies targeting different pathways and may be more consistently achieved with one of those as parenteral prostanoids.^55–57 Evidence is emerging that initial combination of three therapies including a parenteral prostanoid in PAH patients consistently improves functional class, exercise capacity, and survival in relation to a decrease in PVR by 60–70%,^58–60 and reverses RV volumes to within the normal range.⁵⁹ It may be that with the advent of drugs targeting additional signalling pathways, initial quadruple therapy of PAH will bring PVR back to normal or close and normalize RV function.⁶¹

6. Diastolic function

Diastolic function on a PV function loop is defined by an Eed curve. As such, it is a measure of stiffness. More complete definitions of diastolic function include the maximum rate of isovolumic pressure decline (dP/dtmin) and tau (τ), the time constant of isovolumic relaxation.

The Eed curve is more curvilinear than the Ees curve and therefore less adequately described by a linear approximation.^10–12 Diastolic stiffness of the RV is best estimated fitting a non-linear exponential curve through the origin and early and late diastolic PV relationships, with the formula:

where α is a curve fitting constant, and β a diastolic stiffness constant.

The RV diastolic stiffness constant β is a function of ventricular volume and intrinsic changes in myocardial fibre elasticity in relation with titin phosphorylation⁶² and closely associated with disease severity in PH.⁶³ Diastolic stiffness is a cause of increased filling pressures of the RV and backward flow with systemic organ damage.⁶⁴

Diastolic stiffness simply quantified as a one-point end-diastolic P/V ratio retains the prediction capability of more complex β calculation.⁶³ It has been suggested that diastolic stiffness may evolve independently of Ees and emerge as a stronger predictor of outcome.⁶³ Increased diastolic stiffness is an early change in RV adaptation to increased afterload in experimental PH.⁶⁵

Diastolic stiffness is positively related to the ventilatory response to exercise in severe PH⁶⁶ and has been recently reported to be decreased during exercise by β-blocker therapy.⁶⁷

More studies are needed to assess the clinical and prognostic relevance of different indices of RV diastolic function.

7. Ventricular interactions

RV function has also to be understood in the context of diastolic and systolic interactions with LV function.^9,68

A negative diastolic ventricular interaction may occur in PH because of competition for space within the non-distensible pericardium. This alters LV filling and may eventually be the cause of altered LV systolic function and decreased cardiac output. RV volume overload has a more pronounced depressive effect on LV systolic function as assessed by EF than RV pressure overload. This may be explained by differences in systolic ventricular interaction, that is the inotropic effect of LV contraction on RV contraction.^9,68

A positive systolic interaction can be shown experimentally by an improved RV function by an aortic constriction manoeuvre.⁶⁹ In electrically isolated ventricular preparations, LV contraction contributes a significant amount (∼30%) to RV contraction.⁷⁰ Systolic interaction is explained not only by a mechanical entrainment effect of the contraction of shared myocardial fibres but also by LV systolic function determining systemic blood pressure, which is essential to RV coronary perfusion.⁷¹ Increased RV filling pressures and excessive decrease in blood pressure may be a cause of RV ischaemia and decreased contractility.⁷²

An important cause of negative ventricular interaction is regional and inter-ventricular asynchrony with post-systolic contraction or ‘shortening'.⁷³ This develops in parallel to increased PAP and contributes to both altered RV systolic function and LV under-filling.⁷³ Inter-ventricular asynchrony is associated with regional dyssynchrony of RV myocardial fibre contraction. Both RV-LV asynchrony and RV dyssynchrony can be identified and quantified by speckle tracking echocardiography and shown to be associated with a poor outcome in advanced PAH.⁷⁴ However, inhomogeneous regional RV contraction may already be present in early stage or borderline PH⁷⁵ or even in healthy subjects breathing low oxygen mixtures and mild increase in PAP.⁷⁶

8. Stressing the RV-pulmonary circulation unit

Stressing the RV to measure its ‘contractile reserve' defined by systolic function adaptation to afterload may disclose latent failure. The RV may be stressed by an exercise test, fluid loading, or a pharmacologic challenge.

Exercise in normal subjects is associated with an increased Ees and unchanged Ea.⁷⁷ In PAH patients with either normal or decreased Ees/Ea at rest, exercise is associated with a decrease in Ees/Ea.^26,27 This is illustrated in Figure 2. Exercise-induced systolic RV pressure estimated from a tricuspid regurgitation has been shown to be a strong predictor of survival in patients with PAH or CTEPH.⁷⁸ However, a single measurement of tricuspid regurgitation offers an only incomplete description of RV systolic function adaptation to afterload.²⁶ The functional reserve of the RV is ideally measured by a stress-induced increase in Ees. It has been shown that a blunted Ees response to exercise in PH patients is associated with increased RV dimensions and decreased Ees/Ea or EF.⁷⁹ In that study, exercise EF was predictive of outcome with a cut-off value of 37%. A measurement of EF does not require cardiac catheterization and may be adequate to assess RV functional reserve because it is related to Ees/Ea and integrates exercise-induced changes in Ees and RV volumes.⁷⁹ Effects of exercise stress tests on RV PV loops with preserved or exhausted contractile reserve are illustrated in Figure 2.

There has been thought of replacing exercise by a low-dose dobutamine infusion and measuring the RV contractile response by the echocardiographic tricuspid annular plane systolic excursion (TAPSE) or tricuspid annulus S wave.⁸⁰ There is experimental work showing that dobutamine-induced increase in these indices of RV systolic function reflects the resting state of RV-arterial coupling.⁸¹ A dobutamine stress test has also been applied with assessment of the RV contractile response by CMR imaging in congenital cardiac disease. In patients with a tetralogy of Fallot, absence of decreased ESV and increased RVEF in response to low-dose dobutamine has been shown to predict a poor outcome.⁸²

A fluid challenge stresses the RV by a sole increase in venous return. In Starling's heart–lung preparation, an increase in venous return increased RV volumes and SV within a matter of several beats. However, after 3–5 min, EDV returned back to baseline, ESV decreased, and SV remained increased.⁸ Accordingly, in intact human beings, fluid loading may transiently increase EDV, but this is rapidly associated with an increase in Ees, so that Ees/Ea remains unchanged.¹⁰ In patients with severe PH, Ees fails to adapt, resulting in a decreased Ees/Ea.⁸³ This is illustrated in Figure 2.

9. Pathobiology

The pathobiology of the RV response to loading involves changes in cardiomyocyte passive and active tension, a re-emergence of foetal gene expression with an increased expression of natriuretic peptides, a ‘proteomic switch' from α- to β-myosin heavy chain, a metabolic shift, cardiomyocyte hypertrophy and apoptosis, inflammatory cell infiltrates, fibrosis, loss of capillaries, and increased expressions of pro-inflammatory and apoptotic signalling pathways.⁸⁴

In vitro mounted RV cardiomyocytes of idiopathic PAH or congenital cardiac shunt-associated PAH present with an increase in both active and passive tension–length relationships, in relation with decreased titin phosphorylation, and correlated to in vivo Ees and Eed measurements.^62,85 In these studies, myocardial tissue stiffness is explained by a combination of fibrosis and increased myofibril elastance that is partially restored by protein kinase A.⁸⁵ Cardiomyocyte force generating capacity is decreased in SSc-associated PAH compared to idiopathic PAH, in line with in vivo measurements of decreased contractile reserve,²⁷ while diastolic stiffness is not different⁸⁶ (Figure 3). This observation is in keeping with previous suggestion of a dissociation between systolic and diastolic elastances in PAH.⁶²

Evidence has been accumulated in experimental PH models that RV failure is associated with a metabolic remodelling characterized by a switch from oxidative phosphorylation (i.e. generation of energy from the mitochondrial-based fatty acid oxidation) to a glycolytic state (i.e. generation of energy from the cytoplasmic glycolysis).⁸⁷ This metabolic switch has been confirmed in PAH patients along with increased sympathetic tone and insulin resistance and shown to be associated with a toxic lipid accumulation in the myocardium.⁸⁸ These changes were most prominent in patients with loss of function bone morphogenetic protein receptor-2 (BMPR2) mutations, which is typically found in 80% of patients with heritable PAH.^89,90 Mutations of BMPR2 have been reported to limit cardiac hypertrophy and to be linked to defective peroxisome proliferator-activated receptor-γ signalling as well as increased glycolysis, decreased oxidative phosphorylation, and impairment in fatty acid oxidation.^89,90 Presently, it is unknown if interventions aiming at enhancing fatty acid oxidation or improving glucose oxidation might improve RV function.

Microneurographic studies have shown that the sympathetic nervous system is activated in PAH⁹¹ and that this is of prognostic relevance.⁹² A decreased parasympathetic activity in PAH is suggested by impaired post-exercise heart rate recovery and increased RV tissue α-7 nicotinic acetylcholine receptor together with a decreased acetyl-cholinesterase activity.⁹³ This sympatho-vagal imbalance is associated with a down-regulation of RV myocardial β₁ receptors though with an over-expression of β₂ receptors⁹⁴ affecting downstream β-adrenergic signalling and sarcomeric function.⁸⁵ Furthermore, PAH patients also present with an activation of the renin-angiotensin-aldosterone system.⁹⁵ However, interventional or pharmacological neurohumoral modulation strategies studies have not been shown to be effective until now in severe PH.⁹⁶

Survival is shorter in male PAH patients, and this is explained by a better adaptation of RV function to afterload in women.^36,97,98 These differences are essentially explained by differential effects of sex hormones,⁹⁹ but Y chromosome determined BMPR2 signalling might also play a role.¹⁰⁰ Clinical trials of estrogen therapies to prevent RV failure in PH are currently underway.

Patients with idiopathic pulmonary hypertension (IPAH) and a BMPR2 mutation have a worse prognosis than patients in whom this mutation is not found.^101,102 Specific RV abnormalities in calcium handling and sarcomeric function are observed in BMPR2 mutated rats.¹⁰³ The RV of BMPR2 mutated mice presents with altered hypertrophic responses and excessive fibrosis in response to pressure overload.^89,104 Cardiac dysfunction in BMPR2-mutated patients and animals is associated with metabolic derangements and lipotoxicity.^80,89

Inflammation is an inevitable component of the pathophysiology of RV failure on excessive loading,¹⁰⁵ with over-expression of pro-inflammatory and pro-apoptotic mediators with recruitment of macrophages in acute as well as chronic RV failure models in proportion to RV-PA uncoupling.^106,107 A recent study has demonstrated that the nucleotide binding domain, leucine-rich-containing family, pyrin-domain containing 3 (NLRP3) inflammasome in macrophages contributes to RV failure development.¹⁰⁸ The NLRP3 was activated by exposure to damage-associated molecular patterns (DAMPs) resulting in the generation of pro-inflammatory cytokines such as interleukin 1β (IL1β). These findings suggest that RV failure is caused by a vicious circle starting with mitochondrial damage as source of DAMPs to trigger activation of NLRP3 inflammasome, resulting in secretion of pro-inflammatory cytokines such as IL1β and interleukin 6, causing recruitment of monocytes to the right heart.¹⁰⁸

10. The RV in left HF

Left HF is by far the commonest cause of RV failure.¹⁰⁹ Increased pulmonary artery wedge pressure (PAWP) because of increased LV filling pressures or mitral stenosis is initially transmitted upstream to PAP in a close to 1/1 ratio defining an isolated post-capillary pulmonary hypertension (IpcPH). Pulmonary vascular distension sooner or later triggers a remodelling of PA arterioles and capillaries thereby increasing the gradient between PAP and PAWP, defining a combined pre- and post-capillary pulmonary hypertension (CpcPH), formerly called ‘out of proportion' PH.^110,111 The differential diagnosis between IpcPH and CpcPH therefore rests on increased gradients between PAP (mean or diastolic) and PAWP and thus normal or near-normal vs. increased PVR.^109–111 Pulmonary hypertension may complicate HF with reduced LVEF (HFrEF) as well as with preserved LVEF (HFpEF).^109–111 Both IpcPH and CpcPH are associated with a decreased survival rate in HF.¹¹¹

Like in other types of PH, survival of either IpcPH or CpcPH is largely determined by RV function adaptation to increased loading.¹¹¹ A preserved RV EF as measured with radionuclide technology has long been shown to predict exercise capacity and survival in advanced HF.¹¹² Concomitant pulmonary vascular and RV function measurements in HF have shown RVEF to be an independent predictor of outcome independently of PH severity.¹¹³

The failing right heart in patients with HF gives rise to a ‘right heart phenotype', in which symptomatology of systemic congestion predominates.^114,115 The RV and LV are encircled by common fibres and share the septum,^9,68 which is why LV myocardial conditions affect RV systolic function adaptation to afterload. Experimental HFrEF is associated with profound RV-PA uncoupling and very low Ees/Ea in spite of marginally increased PAP.¹¹⁶ Systolic ventricular interactions in HFpEF would be expected to preserve RV function, but these patients also more commonly present with higher PAP on CpcPH, resulting in a disproportional impact of afterload to decrease RV Ees/Ea.¹¹¹ This notion has been confirmed by invasive exercise measurements in HFpEF patients,¹¹⁷ with interestingly a better RV contractile adaptation to exercise-induced increase in pulmonary vascular pressures in women than in men.¹¹⁸

Because of the importance of the RV in HF, simple bedside non-invasive methods for evaluation of RV/PA coupling have been actively developed. The most successful is the ratio of TAPSE (to estimate Ees) to sPAP (to estimate afterload) measured during a standard 2D echocardiography.¹¹⁹ The TAPSE/sPAP ratio is correlated to invasively determined Ees/Ea^120,121 though not always tightly¹¹¹ and sometimes not better than TAPSE alone in HFrEF.¹²² Measurements of RVEF by 3D echocardiography and TAPSE/sPAP are illustrated in Figure 4.

The TAPSE/sPAP ratio has emerged as a strong predictor of outcome in PH on HF,^{110,111,119,122–126} PAH,¹²⁷ or chronic lung conditions.¹²⁸ The prognostic impact of the TAPSE/sPAP ratio on HF is illustrated in Figure 4. The TAPSE/sPAP ratio also predicts outcome of trans-catheter tricuspid,¹²⁹ mitral,¹³⁰ and aortic¹³¹ valve repairs. It is of interest that the TAPSE/sPAP ratio retains its strong prognostic relevance even when PAP is only mildly elevated in chronic obstructive pulmonary disease,¹³² mild uncomplicated systemic hypertension,¹³³ or patients hospitalized with severe acute respiratory syndrome - Corona virus 2 (SARS-CoV2) pneumonia.^134,135

Normal values of the TAPSE/sPAP ratio aggregate around 1.2 mm/mmHg, slightly higher by an average of 0.2 mm/mHg in men than in women and tend to decrease with aging.^136,137 The TAPSE/sPAP ratio decreases with exercise by an average of 0.4 mm/mmHg.^137,138 Rigorously determined prognostic cut-off values for the TAPSE/sPAP ratio vary from 0.31 in patients with severe PH¹²⁷ to 0.64 mm/mmHg in SARS-CoV2 pneumonia with mild increase in PAP.¹³⁴

Other echocardiographic indices of RV-PA coupling, such as the FAC/mPAP ratio, the RV area change/end-systolic area, the TAPSE/PA acceleration time, and the tricuspid S′ wave/sPAP, have attracted less attention than the TAPSE/sPAP ratio and, when evaluated, do not appear significantly correlated to Ees/Ea.¹²¹

There has been interest in assessing the TAPSE/sPAP ratio during an exercise stress test. This was first reported in patients who underwent cardiac resynchronization therapy.¹³⁹ In these patients, the procedure did not affect the TAPSE/sPAP ratio at rest and was associated with a marked improvement of the TAPSE/sPAP ratio during exercise, in correlation to improved exercise capacity.¹³⁹ Exercise TAPSE/sPAP better than resting TAPSE/sPAP may disciminate HFpEF from HF with mild reduction in LVEF.¹⁴⁰

Table 1 provides an overview of indices of RV function, including imaging parameters which may serve as a ‘to do' list for those performing echocardiography or cardiac MRI.

11. RV cardiomyopathies

RV failure may result from direct myocardial insults and/or diseases, even though increased afterloading will at some point occur because of decreased inotropy and heterometric increase in volume.¹⁰

Perhaps, the commonest cause of initial RV cardiomyopathy is ischaemia. Post-mortem examinations¹⁵⁶ as well as imaging studies^157,158 have identified RV necrotic, inflammatory, and oedematous lesions in up to 50% of patients with an acute myocardial infarction. Ischaemic RV cardiomyopathy is more common in inferior infarcts but is also present in a substantial proportion of anterior infarcts.¹⁵⁷ Reported prevalences of decreased RVEF after an acute myocardial infarction range from 20 to 60%.^159,160 However, post-infarction RV dysfunction will most often result from a combination of factors including ischaemia-related decrease in contractility, negative ventricular interactions, mitral regurgitation, and increased pulmonary vascular pressures and is not therefore a good example of initial isolated RV cardiomyopathy with secondary volume overload.¹⁶⁰

Non-ischaemic dilated cardiomyopathy is also associated with decreased RVEF, in one-third to half of the patients, but this is explained by negative ventricular interactions and increased loading in addition to decreased inotropy.¹⁶¹ In hypertrophic cardiomyopathy, CMR studies show a prognostically relevant RV involvement in approximately one-third of the patients.^162,163 Even when standard indices of systolic function are normal, RV free wall deformation, diastolic function, and contractile reserve in response to exercise are reduced in hypertrophic cardiomyopathy.¹⁶⁴

In patients with cardiac amyloidosis, abnormalities in both RV systolic and diastolic function are observed in relation to amyloid deposition but also severity of LV involvement.¹⁶⁵

Acute myocarditis involves the RV in approximately one-fifth of the patients and is then associated with RV dilatation, hypertrophy, decreased EF, and impaired outcome.¹⁶⁶ Altered echocardiographic indices of RV-PA coupling have been reported in COVID-19 patients with no apparent alteration of LV function.^134,135

Systemic diseases such as sarcoidosis¹⁶⁷ or SSc^27,86 may be associated with altered RV function in the absence of overt PH.

A rare but more specifically RV myocardial disease is arrhythmogenic cardiomyopathy.¹⁶⁸ These patients present with a general pattern of progression from a subclinical phase, an electrical phase, and a final structural phase with RV regional or global dilatation and systolic dysfunction.¹⁶⁹ However, there is a lot of overlap and individual variability, as imaging in early stage disease may disclose prognostically relevant abnormal myocardial deformation¹⁷⁰ and increased RV dimensions and dyssynchrony.¹⁷¹ A clinical diagnosis of HF predominantly right sided is made in up to 50% of the patients at time of diagnosis.¹⁷² Gold standard RV-PA coupling measurements have not been reported in arrhythmogenic cardiomyopathy and are therefore of undefined prognostic relevance.¹⁰

12. Conclusions

Recent years have witnessed an improved understanding of the RV in health and disease, with application of rigorous invasive determinations for basic research questions and validation of non-invasive imaging to assess clinical relevance. More clinical research is needed to define optimal RV stress test conditions, but tools are now sufficiently validated for more in depth exploring of the pathobiology of RV failure and testing of therapeutic interventions.

Funding

This study was supported by German Research Foundation, SFB 1213, Project B08.

References

Author notes