Summary
Cardiac resynchronization therapy (CRT) has had a major favourable impact on the care of patients with symptomatic heart failure, left ventricular ejection fraction < 35% and enlarged QRS. Despite this, about 35% of patients who undergo CRT in accordance with current guidelines are “non-responders”. Therefore, more accurate selection of CRT candidates would significantly improve patient benefit and decrease costs. In the past decade, some small non-randomized studies have shown that estimation of left ventricular dyssynchrony by echocardiography might be useful to ameliorate the selection of patients for CRT. These preliminary findings have been challenged by the results of large randomized surveys, such as the Prospect and EchoCRT trials, which demonstrated that no left ventricular mechanical dyssynchrony variable could accurately predict CRT response. In recent years, improvements in myocardial imaging techniques, and the potential of fusion imaging to facilitate our understanding of the physiological basis of dyssynchrony and plan lead delivery, have let us suppose that imaging might play a role in the future of CRT. The aim of the present paper is to provide an overview of recent advances in the field of imaging-guided CRT. The role of imaging in the assessment of CRT candidates, in guiding lead implantation, and in the optimization of CRT delivery will be addressed, together with the limitations of these new techniques.
Résumé
La resynchronization cardiaque par stimulation biventriculaire a fait l’objet d’études démontrant de manière indéniable son intérêt pour mieux traiter les patients insuffisants cardiaques ayant une fraction d’éjection ≤ 35 % et des QRS ≥ 130 ms. Trente-cinq pour cent des patients implantés ne tirent, cependant, pas bénéfice de la thérapie. Il y aurait donc un intérêt à définir des critères plus fins qui permettrait de mieux sélectionner, de mieux implanter les patients susceptibles de tirer bénéfice de la resynchronisation. Les techniques d’imagerie, échographique surtout, ont été testé depuis de nombreuses années avec des résultats globalement insuffisants. Prospect ou Echo-CRT ont semé le trouble sur la valeur de la sélection des patients par l’échocardiographie. Pourtant, il y a eu des progrès dans les techniques et dans la compréhension des asynchronismes. Il y a donc des raisons de penser qu’il pourrait y avoir une place à l’imagerie dans l’optimisation de l’utilisation de la resynchronisation cardiaque dans le traitement des sujets insuffisants cardiaques systoliques. Nous proposons ici une revue des techniques et des utilisations les plus récentes de l’imagerie cardiaque dans la sélection et dans l’aide à l’optimisation du mode de resynchronisation utilisé chez un patient donné. Nous abordons les espoirs et les limites encore existantes dans ces approches.
Background
Cardiac resynchronization therapy (CRT) has had a major favourable impact on the care of patients with symptomatic heart failure, left ventricular (LV) systolic dysfunction and enlarged QRS . Despite this, cumulative evidence shows that 25–35% of patients exhibit an unfavourable clinical or echocardiographic response to CRT .
This evidence has encouraged research into novel indices and strategies that can reliably predict CRT response. In the past decade, multiple single-centre trials and the Cardiac resynchronization in heart failure (CARE-HF) trial have demonstrated that LV mechanical dyssynchrony indices assessed by echocardiography can predict response to CRT and long-term outcome . However, these results were challenged by the multicentre PROSPECT (predictors of response to CRT) trial, in which no single LV mechanical dyssynchrony variable could accurately predict CRT response . Moreover, the EchoCRT trial has shown that in patients with symptomatic heart failure, a narrow QRS complex and echocardiographic evidence of time delays between left ventricle (LV) wall motions, CRT does not reduce morbidity, and even increases mortality . Therefore, the role of imaging in the selection of patients for CRT remains a hot topic of debate.
In this review, we seek to provide an update on the advances in cardiac imaging in the field of CRT, and to discuss the main pathophysiological factors that may determine CRT response. Additionally, we would like to underline the potential role of imaging in guiding lead implantation, and to emphasize the potential benefits of a multimodality approach in the selection of CRT candidates. Finally, we will address the role of cardiac imaging in the optimization of CRT and in the follow-up of CRT recipients.
Role of echocardiography in the selection of CRT candidates
In the normal heart, all LV segments contract in a relative synchronized fashion, and contribute to blood ejection into the aorta. In some patients with heart failure, LV mechanical dyssynchrony may occur, and can be attributed to electromechanical activation delay, regional differences in contractility or regional scars. Many echocardiographic predictors of CRT response have been proposed. Most of these variables are based on measurements of time intervals between velocities of the segments of the LV wall, but their predictive value is not sufficiently robust to replace routine selection criteria for CRT . Also, these time-delay indices are not sufficiently reproducible.
Although QRS duration is currently the main selection criterion for CRT implantation, QRS duration is not perfectly correlated with CRT success rate. Recent European society of cardiology guidelines on heart failure state that CRT is contraindicated in patients with QRS duration < 130 ms . Patients with broad QRS and left bundle branch block (LBBB) but with no mechanical dyssynchrony have been described, indicating that the electromechanical correlation is variable and complex even in the presence of broad LBBB. Nevertheless, until now, it has been believed that only dyssynchrony of electrical origin is likely to respond to CRT . Distinguishing different aetiologies of LV dyssynchrony might explain the substantial fraction of patients with a wide QRS who were non-responders to CRT in large trials; this supports the development of variables and imaging modalities that can facilitate that distinction.
Septal flash and apical rocking: two specific myocardial signatures of mechanical dyssynchrony
The direct mechanical consequences of dyssynchronous contraction induced by LBBB can be described by septal flash and apical rocking. Septal flash is defined as an early fast inward motion of the interventricular septum during the isovolumic contraction period. This motion leads to an outward motion of the anterolateral wall, which pulls the apex laterally during the ejection time, and induces a typical apex motion pattern, described as apical rocking ( Fig. 1 ) . In normal subjects, all LV walls move in a homogeneous fashion towards or away from the apex, depending on the phase of the cardiac cycle (systole versus diastole). In contrast, in patients with dyssynchrony, early activation of one wall, without the corresponding contraction of the contralateral wall, results in the early contracting wall moving away from the original position of the apex, while the non/late-contracting wall is pulled towards this position. When this happens, the displacement of the initially contracting wall will be negative (away from the apex), while the non-contracting wall will be positive (towards the apex), creating the first phase of the rocking motion. Once the late contracting wall is finally activated, it will pull the wall that contracted initially in the opposite direction, creating the second phase of this rocking motion.
Septal flash and apical rocking are qualitative variables that can be easily identified without the use of specific software. These “tools” have been reported as qualitative or quantitative ; both are reported to be highly reproducible and applicable even without any specific expertise. Also, Parsai et al. have shown that the presence of septal flash or its appearance/increase at low-dose dobutamine infusion in CRT candidates is associated with significant reverse LV remodelling after CRT . The value of apical rocking at rest and, especially, during low-dose dobutamine infusion, has been demonstrated, as for septal flash. The presence of apical rocking is associated with both echocardiographic (odds ratio 10.77, 95% confidence interval 4.12–28.13) and clinical (hazard ratio 2.73, 95% confidence interval 1.26–5.91) response to CRT, and the presence of both septal flash and apical rocking in CRT candidates is associated with increased long-term survival . A large retrospective demonstration of the clinical value of apical rocking as a qualitative index has been reported, and a recent paper confirmed its value, and quantified it using a longitudinal strain-derived approach . So, these two approaches are extremely valuable, but are slightly different: septal flash focuses on the isovolumic contraction phase, whereas apical rocking focuses on the systolic phase. They can we both observed in a same patient or distinctively .
Speckle tracking-derived indices: from myocardial signature of mechanical dyssynchrony to cardiac work
Speckle tracking echocardiography (STE) is a relatively new technique that allows the evaluation of LV deformation. Among all strain variables, LV longitudinal strain shows good robustness and reproducibility, which makes it applicable in routine clinical practice . Longitudinal strain assessment is much more reproducible than tissue Doppler, but is also more reproducible than measurements of diameters or pulse Doppler traces, as has been reported . The evaluation of longitudinal strain in different LV segments allows the estimation of residual myocardial contractility, a variable that seems to be strictly linked to CRT response . In dyssynchronous ventricles, delayed segments do not fully contribute to end-systolic function, because their contraction lasts after aortic valve closure ( Fig. 2 ). In the MUSIC study, the strain delay index, which corresponds to the difference between end-systolic strain (measured at aortic valve closure) and peak strain across LV segments, was used to estimate wasted myocardial work, and its entity was significantly correlated to the amount of positive LV remodelling after CRT .
Using quite a similar approach, based on the estimation of strain integrals, Bernard et al. were able to estimate the entity of wasted myocardial work in CRT candidates, and to demonstrate a significant reduction in wasted work in the lateral myocardial wall in CRT responders . Russell et al. recently introduced another non-invasive method to assess regional myocardial work: pressure-strain loops analysis. LV pressure is derived by a normalized reference curve, and the absolute pressure level is estimated from brachial artery cuff pressure ( Fig. 3 ) . The non-invasive LV pressure-strain loop area gives an accurate quantification of regional work, allows the distinction between electrical and mechanical dyssynchrony, and is correlated with regional myocardial glucose metabolism by positron emission tomography . Analysis of regional work during electrical dyssynchrony may provide insights into mechanisms of remodelling and LV dysfunction. Nevertheless, validation data in large clinical trials are still lacking. The key value of these new approaches based on speckle tracking techniques is that they are much more robust than previous approaches and the indices are calculated automatically; this should lead to the best inter- and intracentre reproducibility. The results obtained by some researchers should be investigated and verified in other imaging laboratories and in the field of daily clinical practice.
