Structural, electrical and mechanical remodeling of the canine heart in AV-block and LBBB

Abstract

Long lasting alterations in loading conditions lead to adaptations in structural remodeling (hypertrophy, changes in the size of the cardiac muscle and the cavity and its connective tissue content), mechanical remodeling (abnormal contraction and relaxation of the heart) and electrical remodeling (increased sensitivity to arrhythmias). In canine hearts with atrioventricular (AV)-block (bradycardia-induced volume overload) the three kinds of remodeling follow different time courses, emphasizing the complexity of myocardial adaptation processes. This is further emphasized by the lack of reversibility of electrical remodeling when normalizing volume load by pacing at physiological heart rate. Evidence is increasing that long lasting asynchronous activation (ventricular pacing, left bundle branch block) leads to eccentric hypertrophy with regional differences in structural, electrical and mechanical remodeling.

Introduction

Cardiac resynchronization therapy (CRT) improves pump function of the heart, clinical status and quality of life and reduces hospitalization for heart failure.^1,2 These beneficial effects may be due to some combination of acute improvement of contractile function and reverse remodeling during longer lasting CRT.³ This paper summarizes the causes and consequences of ventricular remodeling, elaborated in detail using data obtained in canine hearts with chronic AV-block and asynchronous activation of the ventricles (due to ventricular pacing and left bundle branch block (LBBB)). Moreover, the reversibility of the various kinds of remodeling is addressed.

Various kinds of remodeling

Cardiac remodeling can be defined as a time-dependent process, involving changes at the level of proteins that represent many functions. Remodeling develops when loading conditions of the heart change, either due to changes of pressure or volume load or due to changes in the contractility of the heart itself (like infarction or asynchronous activation). Important initiators of the remodeling process are stretching of the tissue (myocytes and/or fibroblasts) and elevated levels of circulating hormones, like angiotensin II, endothelin and noradrenalin. Through a large number of pathways, these stimuli lead to various kinds of remodeling: structural remodeling (changes in the size of the cardiac muscle and the cavity and its connective tissue content), mechanical remodeling (abnormal contraction and relaxation of the heart) and electrical remodeling (increased sensitivity to arrhythmias).^4,5

Although it is frequently supposed that electrical and mechanical remodeling coincide with structural remodeling (the hypertrophy or atrophy process), this is not always the case. Moreover, different loading conditions lead to different kinds of hypertrophy. For example, pressure overload (like spontaneous hypertension and aorta banding) leads to concentric hypertrophy as well as electrical remodeling, but depressed contractility may occur only in a late stage.^6–8 AV-block leads to low heart rate and, consequently volume overload, leading to eccentric hypertrophy. Although the degree of structural remodeling is less pronounced than in the previously mentioned models, it is accompanied by mechanical and (pronounced) electrical remodeling.^9,10

Time course of remodeling processes in AV-block

In dogs AV-block was created by radiofrequency ablation and measurements were performed at various time intervals, as has been reported before.^9–13

Structural remodeling starts after 2 weeks of AV-block and progresses for at least 16 weeks approximately, when LV muscle mass has increased by ∼30% and LV cavity volume by ∼60% (Fig. 1). The increase in ventricular wall mass is reached by elongation of cardiomyocytes, rather than through by increasing cell width.^10,14 The phase of most pronounced hypertrophy coincides with the time that plasma levels of neurohormones are increased.⁹ It is, however, unlikely that angiotensin is responsible for the hypertrophy since angiotensin II receptor 1 blockade did not attenuate hypertrophy.¹²

Mechanical remodeling during AV-block is evidenced by changes in the force-frequency relation (FFR). The slope of the FFR was flat at week 0 and negative after 6 and 16 weeks of AV-block (Fig. 2), indicative of a negative FFR. The y-axis (force-axis) intercept increased at week 6, indicating a higher contractile performance at low heart rate. During longer lasting AV-block baseline contractile performance decreased again, as evidenced by a decrease in this intercept from 4315±1458 to 3036±1121 mm Hg/s (⁠

⁠). The increased contractility at low heart rate during chronic AV-block explains why the AV-block hearts were able to maintain cardiac output, after being transiently depressed immediately after AV-block.

In order to illustrate the biphasic character of left ventricular (LV) contractile performance and the progressive structural remodeling in chronic AV-block, LV

was plotted as a function of end-diastolic LV cavity volume (Fig. 3). Acutely after creation of AV-block LV end-diastolic volume increased without an increase in LV

\(\mathrm{d}P/\mathrm{d}t_{\mathrm{max}}\)

⁠. After 6 weeks of AV-block both LV end-diastolic volume and LV

\(\mathrm{d}P/\mathrm{d}t_{\mathrm{max}}\)

increased. Experiments in isolated myocytes, isolated from hearts in this stage of chronic AV-block showed increased intracellular Ca²⁺ levels and NCX current in combination with increased cellular contractility,¹⁴ supporting the idea that after approximately 6 weeks of AV-block the LV is in a hypercontractile stage. This stage disappears during continuing AV-block, as shown by the decrease in intercept of the FFR (Fig. 2) and in LV

\(\mathrm{d}P/\mathrm{d}t_{\mathrm{max}}\)

accompanied by further LV cavity dilation after 16 weeks of AV-block (Fig. 3).

Since abnormalities in the FFR can arise from abnormal calcium handling we measured myocardial protein content of sarcoplasmatic-Ca²⁺-ATPase (SERCA) and Na⁺-Ca²⁺ exchanger (NCX). Absolute contents of both proteins were not significantly changed after 6 and 16 weeks of AV-block, but the myocardial SERCA:NCX protein ratio was significantly lower (−32±14%) after 16 weeks AV-block than in control hearts (Fig. 3). The decrease in the SERCA:NCX ratio was not only accompanied by a decrease in contractility (LV

⁠), but also by changes in relaxation parameters (at baseline and after 16 weeks of AV-block: LV

\(\mathrm{d}P/\mathrm{d}t_{\mathrm{min}}\)

⁠: 2213±390 and 1651±383 mm Hg/s; time constant of relaxation tau 22±5 and 36±10, respectively), indicative of impaired relaxation. Reduced SERCA:NCX ratios are consistently present in failing hearts.¹⁵ The data from the present study indicate a pivotal role of the SERCA:NCX ratio, even in hearts without overt heart failure.

Chronic AV-block also leads to electrical remodeling.^9–11 After onset of AV-block repolarization is prolonged even before development of hypertrophy¹² (Fig. 4). This was measured using QT-time in the surface ECG and the duration of LV and right ventricular (RV) monophasic action potentials, both corrected for heart rate [QT_c-time and APD_c, respectively]. The increase in QT_c-time amounted ∼20% and remained stable up to 16 weeks of AV-block (Fig. 4).¹⁰ The different time courses of structural, mechanical and electrical remodeling (Fig. 4) suggest that these processes may be subject to different regulation processes.

Remodeling due to RV pacing and LBBB

The two basic stimuli for remodeling, stretch and circulating hormones, may also contribute to remodeling during long lasting asynchronous electrical activation. Ventricular pacing and LBBB reduce ventricular pump function, most likely leading to neurohormonal stimulation, as indicated by increased tissue noradrenaline levels in paced hearts.¹⁶ Asynchronous activation also leads to marked regional differences in early systolic stretch and systolic shortening, the highest values of both variables being observed in the latest activated regions.¹⁷

Long-term ventricular pacing leads to ventricular dilatation and asymmetric hypertrophy,^18,19 fiber disarray,²⁰ and altered perfusion distribution.^21,22 The asymmetry of hypertrophy, disturbed perfusion and fiber disarray appear the result of the abnormal contraction patterns and mechanical loading under these conditions, whereas the ventricular dilatation and increased catecholamine concentrations may be related to the reduced pump function.

LBBB patients have a dilated and hypertrophied LV with reduced ejection fraction.^19,23 Preliminary data show that chronic experimental LBBB, as induced by RF ablation in dogs,²⁴ increases LV cavity and wall volume by up to 28±17* and 10±9%*, respectively, within 16 weeks. The ratio of cavity to wall volume (eccentricity index) was similarly increased in LBBB dogs (17±19%*) and LBBB patients (14%*). The ratio of interventricular septum and free wall thickness (asymmetry index) decreased by 0.05±0.05* in LBBB dogs and was 1.00±0.06 and 0.98±0.08* in LBBB and control patients, respectively. However, compared to control LV mass had increased by ∼20% in LBBB dogs and by ∼60% in LBBB patients.²⁵ This difference could be explained by the extensive co-morbidity present in LBBB patients,^19,23 or by the longer duration of LBBB.

A recent study indicates that in asynchronously activated hearts not only structural, but also mechanical and electrical remodeling may be asymmetric.²⁶ This study was performed in dog hearts where heart failure was induced by rapid atrial (synchronous) or ventricular (asynchronous activation) pacing. In the ventricular pacing group, the protein content of proteins involved in calcium handling (SERCA, phospholamban) and impulse conduction (connexin 43) was reduced selectively in the endocardium of the late activated LV free wall.²⁶ All above-mentioned remodeling processes in asynchronous hearts imply a maladaptation of the heart. A vicious circle may originate, where asynchronous activation leads to reduced pump function and asymmetric remodeling and, secondarily dilatation and hypertrophy further reduce LV pump function by mechanical remodeling and by slowing down conduction. Evidence that remodeling due to chronic asynchronous activation may indeed be harmful comes from two studies, showing that heart failure develops sooner in patients who are paced at the RV apex pacing than in patients being either paced at the atrium or only paced in the back-up mode.^27,28

Is remodeling reversible?

Therapies like normalization of arterial hypertension by medical treatment or valve replacement clearly reduce mechanical load on the myocardium. It is, however, less clear whether these interventions also lead to complete reversal of all remodeling processes. To obtain more insight in this issue experiments were performed in dog hearts with chronic AV-block. After 8 weeks of AV-block, the bradycardia-induced volume overload was “treated” by pacing the heart at physiological rate (VDD mode) either at the RV apex or at the RV and LV simultaneously (biventricular pacing) for another 8 weeks. LV cavity and mass significantly decreased within 2 weeks after onset of pacing.^10,29 After 8 weeks of pacing reverse mechanical remodeling was evidenced by normalization of the slope (Fig. 2) and intercept of the FFR.²⁹ However, throughout the 8 weeks of pacing QT_c-time and LV and RV APDc did not show any significant reduction, indicating absence of reverse electrical remodeling.^10,29 A possible role of the abnormal activation sequence, induced by pacing, in the lack of reversal of repolarization parameters seems excluded, because measurements at acute and 8 weeks of pacing were performed both in the presence and absence of pacing, because APD reflects local repolarization processes and because repolarization parameters remained unaltered during the relatively synchronous stimulation in the biventricular pacing group.^10,29

Electrical remodeling in the AV-block model has been associated with an increased NCX current¹⁴ and a decreased I_K current.³⁰ Since reverse contractile remodeling (see above) implies normalization of the NCX current, irreversible changes in the K⁺ channel are likely to be responsible for the absence of reverse electrical remodeling. This suggests that one gene is irreversibly changed, while many other genes, coding for structural and contractile proteins, are reversibly altered.

The finding of poorly reversible electrical remodeling may also apply to other situations. For example, cardiac resynchronization therapy by biventricular pacing often leads to reverse remodeling, evidenced by reduction of LV cavity volume.³ However, reduction of life-threatening arrhythmias, frequently occurring in that particular patient category, has not yet been proven and, based on the present data, should not be assumed on forehand.

References