Electromechanical Dyssynchrony and Resynchronization of the Failing Heart

Abstract

Patients with heart failure and depressed function frequently develop discoordinate contraction due to electrical activation delay. Often termed dyssynchrony, this further depresses systolic function and chamber efficiency, and worsens morbidity and mortality. In the mid-1990s, a pacemaker-based treatment termed cardiac resynchronization therapy (CRT) was developed to restore mechanical synchrony by electrically activating both right and left sides of the heart. It is a major therapeutic advance for the new millennium.. Acute chamber-effects of CRT include increased cardiac output and mechanical efficiency, and reduced mitral regurgitation, while reduction in chamber volumes ensues more chronically. Patient candidates for CRT have a prolonged QRS duration and discoordinate wall-motion, although other factors may also be important as ∼30% of such selected subjects fail to respond to the treatment. In contrast to existing pharmacological inotropes, CRT both acutely and chronically increases cardiac systolic function and work yet it also reduces long-term mortality. Recent studies reveal unique molecular/cellular changes from CRT that may also contribute to this success. Heart failure with dyssynchrony displays depressed myocyte and myofilament function, calcium handling, beta-adrenergic responsiveness, mitochondrial ATP-synthase activity, cell survival signaling, and other changes. CRT reverses many of these abnormalities often by triggering entirely new pathways. In this review, we discuss chamber, circulatory, and basic myocardial effects of dyssynchrony and CRT in the failing heart, and highlight new research aiming to better target and implement CRT as well as leverage its molecular effects.

Introduction

Contraction of the normal left ventricle is geographically coordinated, so that fiber shortening in a given muscle wall layer occurs synchronously and by a similar amount throughout the chamber. This is achieved by the Purkinje conduction system that rapidly delivers endocardial excitation to the myocardium. This contraction pattern can be disrupted by disease in a conducting branch (e.g. a left bundle branch block), or iatrogenically when a ventricle is electrically stimulated at a single site. In both instances, contraction of the left ventricle becomes inefficient, since one side moves inward before the other and is then stretched as the latter continues contracting. This shifts a portion of blood volume back and forth in the chamber rather than ejecting it, resulting in the net decline of systolic performance^{1, 2}.

In the normal heart, the impact of dyssynchrony is quite modest, and for years following its early description, clinical interest was minimal. However, in the 1990’s, investigators began testing whether dyssynchrony might play a more prominent role in the failing heart, and examined pacemaker-like devices that stimulated several places in the heart at once to offset conduction delay. Landmark studies by Cazeau³, Auricchio⁴, Kass^{4, 5}, and others revealed the efficacy of multisite pacing in humans, and led to cardiac resynchronization therapy (CRT), the first use of artificial electrical stimulation for the primary purpose of altering ventricular function.

CRT typically involves electrical pre-excitation of two opposing sites of the heart: the right ventricular septum or apex and the left ventricular free wall (Figure 1). The latter is positioned by retrograde insertion of a pacing electrode through the coronary sinus into a lateral cardiac vein. The two ventricular sites can be stimulated simultaneously or with a slight phase shift (e.g. LV slightly earlier). To assure bi-ventricular stimulation, atrial electrical activation is sensed or artificially paced, and a programmed shorter atrial-ventricular (AV) delay time used to achieve ventricular pre-excitation. The AV delay is set long enough so as not to impair diastolic filling but short enough to achieve bi-ventricular capture. Optimization of the AV delay has a modest impact in most patients, and is often set to a physiological default⁶. Positioning the LV epicardial lead in the lateral wall where activation is most delayed and having myocardium that can conduct this stimulation to the rest of the heart are more important factors for success.

Figure 1.

X-ray of patient receiving cardiac resynchronization therapy. Adapted from Owen et al.¹³⁸.

CRT represents a major advance in heart failure therapy. It is the only clinical treatment that can both acutely and chronically improve cardiac systolic function while also increasing cardiac work and improving long-term survival. Pharmaceutical efforts to improve systole have been useful in the short-term, but so far are detrimental in the long-term. Understanding of CRT first developed from clinical studies dominated by integrative physiology. CRT was viewed largely as a mechanical therapy much like tuning a car engine to produce peak pump performance and fuel economy. This perspective kept the focus on methods to “visualize” dyssynchrony, first using the ECG and subsequently tissue-Doppler wall motion imaging. While there is no doubt that mechanical effects of CRT play a major role in its clinical efficacy, growing evidence supports an extensive and often unique set of molecular and cellular changes that it chronically induces, and these likely play important roles as well. As with other heart failure treatments, CRT ends up impacting a broad range of structural and signaling changes that define the failing heart, and in several instances, does so in manner thus far unique to the treatment. In this context, the previously dyssynchronous failing heart that is resynchronized by CRT does not behave as a heart that was never dyssynchronous to begin with. These differences are suggesting new avenues to leverage the biology of CRT to the broader HF population.

In this review, we examine the chamber pathophysiology of dyssynchrony and CRT in the failing heart, and follow with a discussion of new insights into its more basic cellular and molecular biology. We also highlight current areas of research in the field.

Physiology of Ventricular Dyssynchrony and Resynchronization Therapy

Nearly half of all patients with dilated cardiomyopathy have conduction delays such as a left bundle branch block (LBBB)^{7, 8}. The resulting contractile dyssynchrony generates marked regional heterogeneity of myocardial work⁹ (Figure 2A), with the early stimulated region having reduced load and territories of late activation higher load¹⁰. This is accompanied by matching changes in regional blood flow¹¹. Displacement of blood from early to late and back to early activation sites results in a net decline in ejected stroke volume (Figure 2A)¹². These volume shifts occur when differences in muscle activation are greatest between early and late contracting zones (Figure 2B): during isovolumic contraction and at late-systole into early relaxation (arrows). This is why the rate of pressure rise (dP/dt_max) and late-systolic contraction (shortening after aortic valve closure) are common and sensitive metrics of dyssynchrony. The global effect viewed by pressure-volume relations² reveals a rightward shift of the end-systolic PV relationship (Figure 2C). Both stroke volume (loop width) and stroke work (loop area) decline. In a failing heart, where underlying function is already compromised, dyssynchrony further worsens morbidity and mortality¹³.

Figure 2.

Regional and global pathophysiology of cardiac dyssynchrony. (A) Left panels show strain plots versus time for control dogs and dyssynchronous dogs, in both the early and late activated regions. The late activated region is stretched early in the cycle, and the early activated region is stretched later. Right panels show pressure-strain loops for the early and late activated regions. Adapted from Chakir et al.¹². (B) Time-varying elastance plots for the early and late activated regions, identifying the time of greatest disparity. (C) Pressure volume loops for control and RV (dyssynchronous) pacing. The ESPV relationship shifts rightward in dyssynchrony, indicating reduced function (panels B and C adapted from Helm et al.¹³⁹). (D) Cardiac resynchronization therapy (CRT) has an almost instantaneous effect on function, improving dP/dt within a single beat (adapted from Spragg et al.⁶). (E) CRT also improves the global function, showed via PV loops, reversing the negative changes seen in panel C. (Adapted from Kass et al.⁵) (F) CRT also improves energetics. While dobutamine acts as a positive ionotrope, it does so at a cost: reduced energetic efficiency. CRT improves function and also improves efficiency of contraction (adapted from Nelson et al.¹⁵).

Simultaneous bi-ventricular pre-excitation, the common stimulation mode for CRT, restores coordinate contraction, immediately improving net systolic performance within one beat⁵ (Figure 2D), and augmenting chamber ejection and work (Figure 2E). Importantly, this is achieved without a rise in myocardial oxygen consumption, indicating improved chamber mechanical efficiency^{14, 15} (Figure 2F). Other studies have shown CRT improves coronary blood flow and flow velocity, by enhancing the diastolic wave reflection which causes a backwards-traveling suction-like wave¹⁶. Successfully implemented CRT reduces morbidity and mortality from heart failure ^{17, 18}, and this effect is greatest in patients with basal dyssynchrony^{19, 20}. Indeed, applying CRT to patients without dyssynchrony can be detrimental²¹.

As noted, optimizing CRT timing parameters, atrial-ventricular delay, and RV-LV excitation delay, have a modest impact in most patients^{5, 6, 22}, whereas the site of electrical stimulation has a greater effect. Early stimulation of the RV (as in a LBBB) yields more dyssynchrony and thus LV dysfunction than if the LV is stimulated early (e.g. RBBB)^22–24. This has implications in pediatric patients requiring life-time pacing due to conduction block, and recent studies show function is better preserved if pacing is instituted at the LV apex or mid-lateral wall²⁵. In patients with a LBBB, LV mid-lateral wall pacing alone yields hemodynamic improvements similar to bi-ventricular CRT⁵. While clinical use of LV only pacing has not been subjected to definitive clinical trials, meta-analysis of multiple smaller studies found similar benefits to bi-ventricular CRT²⁶.

CRT also improves functional mitral regurgitation^{17, 27} (FMR) which confers a worsened prognosis in patients with dilated cardiomyopathy. There are several causes of FMR, including chamber/annular dilation that prevents proper coaptation of the valve leaflets, dyssynchrony of papillary muscle contraction, and atrio-ventricular conduction delay that can amplify pre-systolic regurgitation²⁸. CRT reduces FMR by targeting each of these components, since it uses a shortened AV delay to induce ventricular pre-excitation, reduces dyssynchronous contraction, and chronically results in smaller chamber volumes^{29, 30}. Of patients with reduced FMR from CRT, more than 90% have reduced chamber volumes revealing a virtual full overlap between FMR “responders” and CRT “responders”³¹.

Identifying CRT responders

Given its invasive nature and up-front expense, major efforts have been made to identify the patients most likely to benefit from CRT. However, to date, ∼30% of recipients do not show beneficial clinical (or cardiac morphometric) responses, and ongoing efforts to develop more robust predictors continue^{32, 33}. CRT first and foremost alters electrical stimulation, thus an underlying presumption for its use is that the heart has regional activation delay that can be offset by premature excitation. If electromechanical coupling is fairly uniform, then delay in electrical activation generates delay in regional mechanics³⁴. However, heterogeneity of wall stress, contractility, fibrosis, or other abnormalities can also impact electro-mechanical association. Discoordinate contraction itself exacerbates electro-mechanical delay in the late activated (high stress) regions^{35, 36}. Furthermore, wall motion can appear dyssynchronous even when electrical activation is normal, due to regional heterogeneity of myocardium function³⁷. For example, right-left heart dyssynchrony occurs with pulmonary hypertension^38–40 without QRS widening⁴⁰ thought due to higher RV wall stress, and CRT can reverse this⁴¹. Disparities in wall contraction timing following myocardial infarction can appear as dyssynchronous⁴², though this is far less likely to be ameliorated by CRT.

Current guidelines identify CRT as a class I recommendation for patients with a QRS complex >150 ms (and EF< 35%), and a class IIa recommendation for patients with a QRS complex between >120 and <150 ms^{43, 44}. However, meta-analyses of 15 large CRT clinical trials found QRS duration predicts 2/3 of responders⁴⁵, and other studies found no correlation between QRS duration and various parameters of functional or cardiac remodeling outcome^{46, 47}. Alternative measures of dyssynchrony have focused on regional wall motion. Early excitement over tissue-Doppler methods^{48, 49} was dampened by the multicenter PROSPECT trial⁵⁰ and revealed marked variability in dyssynchrony index assessment even between expert cores, and a poor overall predictive utility. Newer methods involve speckle tracking^51–54, 3-dimensional echocardiography^55–58, MRI-based techniques^59–61, and SPECT nuclear imaging^{62, 63}. Some integrate information regarding both the magnitude and spatial distribution of dyssynchrony²³. Still, definitive trials remain needed if any of these approaches are to become widely accepted and used^{32, 33}.

A drawback to primarily focusing on myocardial motion to index dyssynchrony is that it ignores the corresponding electrical activation that is central to the impact of CRT. For example, mechanical dispersion in the absence of electrical delays occurs in patients with a narrow QRS complex. A role for CRT in this setting was suggested early on^{64, 65}, however, more recent multi-center studies did not show benefit^{66, 67}, with one terminated prematurely due to futility and safety concerns⁶⁷. Methods combining electrical and mechanical assessments have been reported^{68, 69}, but remain to be rigorously tested in clinical trials. Lastly, other types of biomarkers, such as blood born or heart tissue assays may ultimately be found to define a responsive heart.

Does CRT involve more than LV Mechanics?

While a guiding theory of CRT has been that its efficacy relates principally to improved cardiac mechanics, persistent difficulties in predicting responsive patients suggests something is being missed. Basal dyssynchrony⁷⁰ and acute resynchronization of wall motion after CRT implantation⁷¹ have both been reported to predict responder patients. However, the relation between chronic response magnitude and such wall motion measures is weak. For example, Bleeker et al.⁷¹ found CRT must acutely reverse LV dyssynchrony to lead to chronic reduction in end-systolic volume (reverse remodeling index). However, the extent of resynchronization among these patients, all with Class I indications for CRT, varied markedly. Beyond identifying a minimal threshold of response required to observe any benefit, the data showed no correlation between the magnitude of re-coordinated wall motion and chronic ESV decline (Figure 3). Similar issues apply to recent work using radial strain delay analysis⁷⁰ where a threshold effect linking chronic response to basal dyssynchrony exists, but beyond this, the association is weak. Other approaches such as 2-D echo-cardiographic speckle tracking ⁷² to optimize lead placement or cardiac output to optimize AV and VV delays⁷³ have been tested, and while both have been suggested to lower the non-responder rate, it still remains near 30%.

Figure 3.

The relationship between immediate LV resynchronization, and reduction in LV end-systolic volume at the 6-month follow-up from Bleeker et al.⁷¹. While an acute response to CRT (>20% LV resynchronization) appears necessary for a positive chronic response (>10% LV reverse remodelling), among these “responders” the relationship between acute and chronic response is very poor (see inset). Thus, acute hemodynamic benefit to CRT has little ability to predict long-term benefit. This suggests there is more to dyssynchrony and CRT than LV mechanics.

A number of factors likely contribute to a persistent 30% non-responder rate among patients clinically indicated to receive CRT. Some have poor venous anatomy and difficulty in advancing leads in the appropriate territory limits effectiveness. In others, the underlying disease maybe too advanced, or the underlying myocardium too heterogeneous due to scar tissue that impedes effective electrical activation from the pacing lead. Another cause is that the assessment of dyssynchrony was misleading – in that wall motion discoordination was not really due to electrical conduction delay but rather to regional myocardial properties, and thus hard to offset by pre-excitation.

Lastly, we should entertain the hypothesis that equating CRT efficacy solely to its impact on regional mechanics and hemodynamics is itself limiting. Like all heart failure therapies, secondary changes in molecular and cellular function are also induced, and over time, these become intertwined with the mechanical effects altering the pathobiology. It would be difficult to dissociate these mechanisms, yet molecular/cellular changes induced by dyssynchrony and/or offset by CRT pose a novel perspective for the field. For one, such changes may pave a path to alternative biomarkers for responders. Secondly, molecular responses induced by CRT might themselves be leveraged as heart failure therapy in patients that are not dyssynchronous. Again, the unique feature of CRT is that systole is acutely and chronically improved yet mortality is also enhanced. This is not simply a matter of improving cardiac output but how it is done at the organ, cellular, and sub-cellular levels. In the next section, we review studies revealing such mechanisms.

Cellular and Molecular Aspects of Dyssynchrony and Resynchronization

Advances in the basic science of dyssynchrony and resynchronization began taking shape after the therapy was clinically approved for human use in the early 2000’s. Spragg et al.⁷⁴ reported the first data showing dyssynchrony induced regional enhancement of stress kinase expression and depression of calcium handling proteins in the failing canine heart. This animal model was then modified to include 6-weeks tachypacing with dyssynchrony (left bundle ablation) or 3 weeks dyssynchrony followed by 3-weeks bi-ventricular tachypacing (CRT). The latter displayed a slightly improved EF over dyssynchronous heart failure⁷⁵; however, both models manifested dilated hearts with depressed global systolic function and elevated diastolic pressures. Other variants of this model have used longer pacing periods ranging 4–24 months^{76, 77}, though these pose some practical (and expense) limitations.

Effects of Dyssynchrony and CRT on Myocyte Function and Ca²⁺ handling

Myocyte calcium handling is rendered abnormal by heart failure with or without dyssynchrony, but recently studies indicate that restoring synchrony has benefits even in the absence of much improvement in global chamber function. Sarcomere shortening in the dyssynchronous failing heart declines and contraction and relaxation kinetics are slowed¹², and both are coupled to reductions in whole-cell calcium transient amplitude and delayed dynamics^{12, 77–80} (Figure 4A). These changes were observed similarly in both early and late activated territories in HF with dyssynchrony, and in cells from epicardium versus endocardium¹². By contrast, the L-type calcium current was preserved in early-activated myocardium but reduced in the lateral wall⁷⁹ (Figure 4B). CRT substantially reversed nearly all of these changes^{12, 77, 79}. These cellular improvements occurred despite depressed global function accompanying persistent tachypacing, revealing a favorable impact of resynchronization per se. Improved calcium cycling from CRT was not purely related to synchrony but required a prior history of dyssynchrony, as cells from failing hearts that were never dyssynchronous also displayed abnormal behavior similar to those from dyssynchronous hearts⁷⁸.

Figure 4.

Dyssynchrony reduced baseline cellular function, which is restored by CRT. (A) Myocyte sarcomere shortening and corresponding whole cell calcium transients in myocytes taken from control (Con), dyssynchronous heart failure (HF_dys), synchronous heart failure (HF_sync), and CRT. Adapted from Chakir et al.⁷⁸. (B) Peak current-voltage relationships for I_Ca,L. (C) Western blots for calcium handling proteins. Arrows show the direction of change in HF_dys and CRT groups as compared to control. Panel B and C adapted from Aiba et al.⁷⁹. (D) Three-dimensional reconstructions from confocal microscopic images showing t-tubule structure (blue) and ryanodine receptors (red). Adapted from Sachse et al.⁸⁰.

The mechanisms for altered basal function and calcium handling by CRT remain to be fully defined. Human biopsy data has reported upregulation of phospholamban (PLN)⁸¹ and Serca2A⁸² in CRT responsive patients. In the canine model, Aiba et al.⁷⁹ found PLN and Serca2A protein expression declined and Na⁺-Ca²⁺ exchanger levels rose throughout the myocardium in HF with dyssynchrony (Figure 4C). However, CRT did not reverse these changes and/or regional modifications were often at odds with Ca²⁺-transient changes in the same territory. Alternative mechanisms such as post-translational modifications (phosphorylation, oxidation, or nitrosylation) or structural changes are still being investigated. Another mechanism is a change in the structural organization of transverse t-tubules and the sarcoplasmic reticulum reported by Sachse et al.⁸⁰. Dyssynchrony disrupted the regular longitudinal spacing of the t-tubular system and its registration with ryanodine receptors (Figure 4D) in the lateral but not anterior wall, and this was partially reversed with CRT. Structural disruption of the tubular-SR has been reported in failing hearts and thought to contribute to abnormal calcium handling⁸³. Lastly, the responsiveness of the myofilaments to calcium also appears to be impacted by dyssynchrony and CRT. Preliminary data reveals significantly lower Ca²⁺ sensitivity and depressed maximal activated force in dyssynchronous HF this is restored to normal after CRT⁸⁴. Work to identify the mechanism(s) is ongoing, but new quantitative mass-spectrometry methods may pave the way for assessing post-translational changes in myofilament proteins that control Ca²⁺ sensitivity⁸⁵.

CRT and beta-adrenergic signaling

CRT responsive patients display an enhanced cardiac responsiveness to sympathetic stimulation^{86, 87}. Acutely, CRT blunts efferent sympathetic tone that is often elevated in heart failure patients⁸⁸, and chronically CRT results in upregulation of β1-adrenertic receptor gene expression in responsive patients⁸⁹. More detailed analysis of β-adrenergic signaling has been performed in the canine model. Myocytes from always dyssynchronous or synchronous HF display a blunted response to isoproterenol, affecting both sarcomere shortening and calcium transients (Figure 5A). Both responses are restored to near normal with CRT⁷⁸, accompanied by increased β1-receptor expression and membrane density, and adenylate cyclase activity¹².

Figure 5.

Beta-adrenergic pathways are reduced in HF_dys, but restored via RGS2/3 signaling in CRT. (A) In response to isoproterenol, HF_dys exhibited a much smaller augmentation in sarcomere shortening and calcium transient amplitude, which was reversed by CRT. (B) Pertussis toxin (PTX) preferentially inhibits Gαi signaling, and it restored the isoproterenol response in HF_dys, but had no effect on CRT, suggesting that CRT enhances β-adrenergic response via this pathway. (C) RGS2/3 can inhibit Gαi, and were significantly up-regulated in CRT. (D) Example tracings of SR-AKAR3 FRET in zinterol-treated cells. PKA activation in the SR was only observed in the CRT model. Adapted from Chakir et al.⁷⁸.

Experimental studies also revealed a unique effect from CRT on β2-adrenergic signaling via coupling to inhibitory G-protein. Gαi signaling was increased in always synchronous or dyssynchronous HF, as observed in human heart failure^{90, 91}, and blunted cAMP activation and downstream activation of protein kinase A at the sarcoplasmic reticulum (Figure 5D). Incubation of cells with pertussis toxin (PTX), which blocks Gαi, increased these responses upon exposure to zinterol, a β2-agonist (Figure 5B). In contrast, cells from CRT hearts had a greater basal zinterol response that was unaltered by PTX. Gαi protein expression was elevated in each of these models similarly, but CRT enhanced co-expression of two negative regulators of G-protein signaling, RGS2 and RGS3 that can suppress Gαi-signaling⁷⁸ (Figure 5C). Upregulation of either RGS protein in myocytes from dyssynchronous or synchronous HF resulted in behavior similar to that of myocytes from CRT hearts⁷⁸, and were also no longer responsive to PTX. RGS2 and RGS3 gene expression were also increased in human LV biopsies from CRT responders, providing translational support for this mechanism⁷⁸.

CRT modulation of ion-channel expression and function

Ventricular arrhythmia is a common complication and cause of death from HF. Many channelopathies are involved which often lead to prolongation of the action potential duration (APD) and slowed conduction velocity. Pharmacological targeting of these changes has been generally counterproductive, and defibrillation by implantable devices serves as the primary therapy. In HF patients receiving CRT, ventricular arrhythmia declines^92–96, and while some raised concerns that epicardial stimulation used by standard CRT may be pro-arrhythmic⁹⁷, but in practice, CRT appears to be anti-arrhythmic⁹⁸, reducing the incidence of a first ventricular tachycardia event⁹⁹. This is another unique feature of CRT, in that it directly enhances systolic function while also countering malignant arrhythmia.

Mechanisms for dyssynchrony and CRT-related electrophysiological changes have been explored in some detail in canine models. Dyssynchrony without LV dysfunction can itself induce regional abnormalities of conduction and repolarization⁷⁴. Whereas conduction is normally fastest in the endocardium, in non-failing hearts with a left bundle branch block, the epicardium became fastest in the late activated lateral wall. This was accompanied by a dislocalization of connexin 43 (Cx43) from the intercalated disk to the longitudinal plasma membrane. The latter was not observed, however, in a long-term study of dyssynchrony only in piglets¹⁰⁰.

Heart failure involves many channel abnormalities^101–103 which may contribute to conduction block¹⁰⁴, and when both are combined, one observes substantial prolongation of action potential duration (APD) associated with declines in multiple ion currents, and worsened conduction¹⁰⁵. APD is prolonged more in the late contracting zone in dyssynchronous hearts⁷⁹, and CRT partially reverses this^{81, 83}. Specific repolarizing potassium currents, such as the inward rectifier K+ current (I_K1), transient outward K+ current (I_to), and delayed rectifier K+ current (I_K), decline in dyssynchronous HF⁷⁹, concordant with a fall in protein expression for the corresponding channel proteins (Kir 2.1, Kv4.3 and KChIP2, and KvLQT1, respectively). CRT partially reversed changes in I_K1, and I_K (but not I_to) and their related proteins. CRT also reverses increased late sodium current (I_Na-L) observed in dyssynchronous HF. The I_Na-L inhibitor, ranolizine, shortened the APD and reduced the incidence of early afterdepolarizations (EADs) in dyssynchronous HF but had no impact on either behavior in myocytes from CRT dogs. The combined effect of CRT on these repolarizing currents reduces arrhythmia. It is again important to remember that these electrophysiological changes from CRT are being observed in an animal model in which chamber dilation and dysfunction persists, and so is more likely reflecting the impact of discoordinate contraction that is resynchronized.

Modulation of cell survival signaling

Myocyte apoptosis contributes to cardiac dysfunction. It is observed in dyssynchronous heart failure in both experimental models and humans. In humans, markers of apoptosis were elevated in LV biopsies from dyssynchronous HF and subsequently reversed after initiating CRT¹⁰⁶. Another study found the pro-apoptotic protein annexin A5^{107, 108} declined in patients responding to CRT¹⁰⁹. Anti-apoptotic effects of resynchronization has been observed in canine⁷⁵ and pig models¹¹⁰.

Several stress kinases that can trigger apoptosis¹¹¹ are also more active in dyssynchronous HF; including septal ERK1/2 MAPK expression, and increased lateral wall activity of p38 MAPK, CaMKII, and TNF-α⁷⁵. However, these changes are regional, appearing in the late activated high stress wall, whereas apoptosis is more globally impacted, indicating other modulators are involved. One kinase that is globally inactivated is Akt that is in turn linked to reduced BAD phosphorylation and consequent pro-apoptotic signaling. CRT reverses both the regional stress kinase and global Akt activation changes.

Biomarkers of CRT in circulating blood

While one might not anticipate that discoordination in electromechanical timing of the heart might lead to detectable circulating plasma markers, a number of clinical studies have explored this possibility and reported limited but positive findings. Examples include matrix modulators such as protein tenascin-C, and metallo-proteinases 2 and 9 (fall after CRT¹¹²), tissue inhibitor of MMPs (TIMP-1, rises after CRT¹¹³), and the G-coupled receptor ligand apelin^{114, 115} (rises with CRT). Whether these are specific to dyssynchrony or reflect overall enhanced LV function by CRT is uncertain.

CRT and mitochondria energetics

Heart failure is often considered a disease involving energy starvation^{116, 117}. The dyssynchronous heart is inefficient at the chamber level, and regional disparities in metabolism reflected by glucose uptake have been documented^{118, 119}. These, along with global chamber efficiency, are improved by CRT^{15, 120} (e.g. Figure 2F). However, recent evidence shows CRT also favorably impacts underlying mitochondrial function. Mitochondrial basal oxygen consumption is increased in canine dyssynchronous heart failure¹²¹ yet accompanied by a decline in ATPase activity¹²². CRT increased the mitochondrial respiratory control ratio, an index of ATP synthetic capacity (Figure 6A), to levels similar to those in healthy controls. CRT altered expression of at least 31 different proteins within the mitochondrial proteome, many involving the respiratory chain¹²¹.

Figure 6.

Effect of dyssynchrony and cardiac resynchronization on mitochondrial function and oxidation. A) Respiratory control ratio (mitochondrial O₂ consumption in present of substrate +ADP versus –ADP) is depressed in dyssynchronous heart failure (HF_dys) and augmented to sham control levels by CRT. * p<0.001 vs CRT, p<0.05 vs Con; # p<0.05 vs CRT. B) ATP synthase activity (measured by in gel assay) at baseline and with exposure to reducing agent dithiothreitol (DTT). Activity was reduced in HF_dys versus CRT but augmented to control levels by DTT. This indicates oxidation plays an important role in depressed mitochondrial function in HF_dys. Adapted from Agnetti et al.¹²¹ and Wang et al.¹²².

Improved oxidative phosphorylation following CRT has been further mechanistically linked to oxidative post-translational modifications in the α-subunit of ATP-synthase¹²². Upon exposure to reducing conditions, ATPase activity in dyssynchronous HF reached levels in control or CRT myocardium, supporting an altered oxidative environment in dyssynchronous HF (Figure 6B). Two disulfide bonds between ATP synthase-α subunits were increased in dyssynchronous HF (Cys 294-Cys294 and Cys294-Cys103) and reduced by CRT. Cys294 was also S-glutathionylated, and CRT suppressed this while favoring S-nitrosylation at the same residue. S-nitrosylation of ATP-synthase α occurs in ischemic preconditioning, and may be cardioprotective¹²³. Thus, the S-nitrosylation /S-glutathiolylation balance of appears to be a regulatory mechanism for ATP synthase that is targeted by CRT.

Figure 7 summarizes the various molecular and cellular changes that have been identified in both heart failure with dyssynchrony and the response to CRT.

Figure 7.

Summary of molecular alterations in dyssynchronous heart failure and CRT. (Illustration Credit: Ben Smith).

Clinical Challenges and Future Directions

Lead placement and optimal implementation of CRT

While the LV lead used in CRT is generally placed in the lateral and/or postero-lateral wall, this can vary among patients and ongoing efforts continue to improve on optimizing this placement. Several studies indicate the LV free-wall territory in which effective CRT can be obtained is sizable (∼40% of the free wall^{124, 125}), but coronary venous anatomy may preclude even this lead placement¹²⁶. New leads with multiple selectable electrodes may improve on this limitation and reduce the non-responder rate¹²⁷. Others are exploring LV endocardial pacing either via transseptal puncture¹²⁸, or a wireless system with micro-stimulators placed within the LV endocardium that are actuated by ultrasound to stimulate the muscle^{129, 130}. These remain works in progress.

CRT and Atrial fibrillation

Atrial fibrillation (Afib) is common in moderate to severe heart failure patients, and it makes implementation of CRT more difficult since timing for pre-excitation of the LV cannot be accurately predicted¹³¹. In the Resynchronization for Ambulatory Heart Failure Trail (RAFT), patients with permanent Afib were randomized to a defibrillator or defibrillator plus CRT, and CRT showed minimal benefits, with a trend towards reduced HF hospitalizations¹³². While CRT superimposed with atrio-ventricular node ablation guarantees bi-ventricular stimulation, one compromises physiological heart rate regulation. This remains a population where the efficacy of CRT is compromised.

Might transient dyssynchrony be beneficial?

While sustained dyssynchrony has been well established to worsen HF outcomes, several studies have explored the impact of brief exposure to dyssynchrony where the effect is quite different. Prinzen and colleagues subjected rabbit hearts to short ventricular pacing (dyssynchrony) prior or following coronary occlusion, and observed major reductions in infarct size^{133, 134}. Post-conditioning efficacy of transient dyssynchrony has also been reported in pigs¹³⁵. Two factors involved with ischemic post-conditioning, HIF-1α and HSP-70¹³⁶, are induced by dyssynchrony¹³⁷, with activation observed in the region opposite to where pacing was instituted (e.g. late contracting wall), suggesting a role for increased wall stress.

Another example where short-term exposure dyssynchrony was beneficial was in the failing heart. In dogs subjected to 6-weeks of atrial tachypacing (synchronous HF), 2-weeks of RV tachypacing (dyssynchrony) was substituted during week 3 and 4. Myocytes from the latter hearts displayed substantial improvement in rest and β2-stimulated myocyte function and calcium transients⁷⁸, and a decline in Gαi-coupling.

The molecular consequences from transient dyssynchrony remain to be determined, but these recent studies raise intriguing possibilities. For example, might transient dyssynchrony in an otherwise synchronous failing heart improve myocardial function and/or cytoprotection? Should CRT patients undergo occasional pacing holidays (restoring dyssynchrony from the underlying conduction block) perhaps re-igniting the biological response to resynchronization? Can transient dyssynchrony be used to offset ischemic damage in the human heart? These and may other related questions will need to be explored in future research.

Conclusions

CRT remains a major advance in HF therapy since the adaption of beta-blockade and implantable defibrillators. While understood and accepted for its capacity to improve chamber level mechanics and thus circulatory hemodynamics, chronic CRT has revealed myocardial changes that were not previously envisioned. Whether this is all attributable to improved global function may be impossible to unravel, though the capacity to reverse abnormalities often to healthy control levels, as well as some of the more unique changes observed, suggests more is at play. More definitive analysis of the importance of specific molecular changes due to dyssynchrony and/or CRT will require the use of genetically engineered models, and this work is ongoing. Nonetheless, for a treatment that first sounded to many to be too simple to achieve much benefit, CRT has proven the opposite on both counts.

Non-Standard Abbreviations/Acronyms

Afib

Atrial Fibrillation

APD

Action potential duration

AV

Atrial-ventricular

CRT

Cardiac Resynchronization Therapy

dP/dt_max

Maximum rate of pressure rise

EAD

Early afterdepolarizations

ECG

Electrocardiogram

EF

Ejection fraction

FMR

Functional Mitral Regurgitation

HF

Heart Failure

HF_dys

Dyssynchronous heart failure

LBBB

Left bundle branch block

LV

Left Ventricle

MMP

Matrix metalloproteinases

PLN

Phospholamban

PTX

Pertussis Toxin

RBBB

Right bundle branch block

RGS

Regulators of G-protein signaling

RV

Right ventricle

TIMP

Tissue Inhibitors of Matrix Metalloproteinases

VV

Ventricular-ventricular