This article discusses how echocardiography can be applied to quantify dyssynchrony in patients who are evaluated for cardiac resynchronization therapy (CRT). A number of echocardiographic indices have been proposed as markers of success of CRT. However, when tested against QRS width in prospective clinical trials, none of the echocardiographic indices are proven to give clinical benefit. One important message in this review is that future studies should focus on approaches which can differentiate between electrical and non-electrical aetiologies of dyssynchrony, since only electrical dyssynchrony is likely to respond to CRT. Just measuring velocity indices does not identify the aetiology. Myocardial strain appears more promising, but one should be aware that timing of peak systolic strain is determined not only by electrical conduction. It is proposed to use onset septal shortening during pre-ejection for timing of earliest left ventricular (LV) electrical activation. One should take into account potential ischaemia, scarring, and other structural changes as contributors to dyssynchrony. As a method to identify electrical dyssynchrony, the authors propose to use time of active force generation as defined by LV pressure-strain loops. A non-invasive method to measure segmental pressure-strain loops is also proposed as a means to quantify the impact of dyssynchrony on distribution of myocardial work. Furthermore, it is important to be aware that LV dyssynchrony may have a combination of aetiologies, not all amenable for CRT.
Cardiac resynchronization therapy
In patients with heart failure and left ventricular (LV) intraventricular dyssynchrony, cardiac function can be improved by electrically activating the septum and the lateral free wall in a synchronized manner with a biventricular pacing device, a treatment named cardiac resynchronization therapy (CRT). In selected patients with congestive heart failure, CRT causes reversal of LV remodelling, improvement of symptoms, and reduction in mortality.1,2 Furthermore, CRT tends to reduce mitral regurgitation.3 However, about one-third of patients treated with CRT do not show clinical improvement, indicating that we need better methods for identifying those who will benefit from the therapy.4
Identification of CRT responders can be done by two different principles, i.e. by assessing either electrocardiographic or cardiac mechanical signs of dyssynchrony. At the present time, electrocardiographic criteria are used as the main screening tools, and patients with congestive heart failure and ejection fraction <35% are selected for CRT based on morphology and duration of the QRS complex. In clinical trials that have documented clinical benefit of CRT, one of the entry criteria has been QRS duration over 120 ms, and average QRS duration in the large trials has been over 150 ms, most often with left bundle branch block (LBBB) type morphology and sinus rhythm. There is not sufficient documentation to recommend CRT in patients with pure right bundle branch block.5–7 Several trials have shown that QRS is not a very precise marker of LV dyssynchrony, as patients with a wide QRS may not have LV dyssynchrony and those with a narrow QRS may actually have dyssynchrony.8–10 This may explain why a substantial fraction of patients with wide QRS are non-responders to CRT in the large trials.9,11
A number of echocardiographic indices have been proposed as markers of success of CRT.12 However, when tested against QRS width in prospective clinical trials, none of the echocardiographic indices are proven to give clinical benefit in patient selection for CRT.13,14 Although QRS duration is an imperfect marker of responders to CRT, there is yet no other consensus definition of dyssynchrony that has been sufficiently evaluated. Therefore, current international guidelines do not recommend routine quantitation of dyssynchrony by echocardiography in the evaluation of patients who are considered potential candidates for CRT.1,5,7 The value of using QRS width as a selection criterion is best documented in patients with QRS width ≥150 ms who have the greatest benefit from CRT.5,15–17 In patients with a QRS width of 120–150, results from CRT are more variable.5,18
Assessment of inter-ventricular dyssynchrony
Inter-ventricular dyssynchrony is measured by Doppler echocardiography as the difference between LV and right ventricular pre-ejection intervals calculated from the onset of QRS on ECG to the onset of aortic outflow and pulmonary outflow, respectively. It has been shown that reversed LV remodelling and improved LV function during CRT is associated with a reduction in inter-ventricular dyssynchrony.19 However, most evidence suggests that inter-ventricular dyssynchrony is not as useful in the prediction of response to CRT as LV intra-ventricular dyssynchrony.12
Assessment of left ventricular intra-ventricular dyssynchrony
Two-dimensional and M-mode echocardiography
The initial step in the assessment of intra-ventricular dyssynchrony should be conventional grey-scale two-dimensional (2D) imaging and one should look for abnormal septal motion. Objective evidence of abnormal septal motion can be obtained by M-mode echocardiography which in typical cases shows rapid pre-ejection leftward motion of the septum, followed by rightward motion, a pattern which is named septal beaking.20 This phenomenon can also be observed by tissue Doppler and is then named ‘septal flash’ (Figure 1).21 Although present only in some patients with LBBB, septal flash has shown to be a good predictor of response to CRT21 both in terms of clinical response and reverse remodelling. Another abnormal motion pattern which is consistent with significant LV intraventricular dyssynchrony is apical rocking.22 In addition to these qualitative methods, a number of quantitative echocardiographic measures of dyssynchrony are available.
Tissue Doppler colour M-mode showing septal flash which represents early-systolic rapid septal thickening and thinning, in a patient with left bundle branch block. The septal flash is resolved post-cardiac resynchronization therapy. Modified from Parsai et al.21
The technically simplest approach to quantify LV dyssynchrony is septal-to-posterior wall motion delay by conventional M-mode echocardiography in the parasternal short-axis view.23,24 This method may work in patients with non-ischaemic cardiomyopathy, but is more difficult to apply in ischaemic cardiomyopathy, since previous infarcts and akinesia in the measurement regions make it difficult to define the peaks of septal and posterior wall inward motion.
Tissue Doppler velocity imaging and speckle-tracking strain imaging
Velocity imaging by tissue Doppler and strain imaging by speckle-tracking echocardiography (STE) have been the preferred echocardiographic methods to quantify mechanical dyssynchrony. A detailed description of the procedure is found in a consensus statement from the American Society of Echocardiography.12 One approach is to assess velocities in the basal part of six basal and six mid-LV segments, and to measure time from beginning of the QRS signal to ‘peak’ systolic velocity during LV ejection25,26 (Figure 2). Regional differences in time to peak systolic velocity can be expressed as standard deviation for measurements from all 12 segments and has been shown to predict reverse remodelling in a number of smaller studies with CRT.26,27 Maximal delay in peak velocity between the anterior, inferior, septal, and lateral walls has also been shown to predict clinical response to CRT and reverse remodelling.25,26 As discussed later in this article, the principle of using ejection-phase velocities to identify candidates for CRT may not be optimal and may even be misleading. A recent article addressed this limitation of ejection-phase velocities.28 The issue about validity of using ejection-phase velocities is essential to resolve since several trials, including the negative PROSPECT trial,13 used myocardial ejection velocities to quantify dyssynchrony. Recent studies suggest that timing of peak strain may be a more useful measure of dyssynchrony when evaluating patients for CRT.29–31 A study which tested most of the proposed echocardiographic indices, however, concluded that none of the current indices predicted clinical response to CRT.14
Colour-coded tissue Doppler study from three standard apical views of a patient who responded to resynchronization therapy. Pre-treatment time–velocity curves from representative basal or mid-levels are shown. Maximum opposing wall delay (of peak ejection velocity) was seen in an apical long-axis view of 140 ms between the septum and posterior wall, consistent with significant dyssynchrony (≥65 ms). AVC = aortic valve closure; AVO = aortic valve opening. Gorcsan et al.12
As discussed in the subsequent paragraphs, one reason for the apparent failure of echocardiography to provide added value in selection of patients for CRT may be that suboptimal methodological approaches have been used. In particular, lack of methodology to differentiate between electrical and non-electrical aetiologies of dyssynchrony is important since it is primarily dyssynchrony caused by electrical conduction delay which will respond to CRT. Furthermore, the use of ejection-phase velocity indices may have confused the issue because peak ejection velocity has significant limitations as a marker of electrical conduction delay. In addition, there is a need for better standardization of measurement techniques. Therefore, it may be too early to conclude about the diagnostic value of echocardiography in selection of candidates for CRT.
In the following paragraphs, we will take a critical look at some of the principles behind current echocardiographic methods and we propose new diagnostic approaches which may be worth exploring.
Differentiation between electrical and non-electrical causes of dyssynchrony
Dyssynchrony is defined as uncoordinated regional myocardial contractions32 and may in principle have three different aetiologies: (i) electrical conduction delay which causes non-uniform timing of myocyte depolarization (i.e. ‘primary electrical dyssynchrony’), (ii) abnormalities in excitation–contraction coupling, and (iii) abnormal myocardial contractility or load which cause regional delay in onset shortening and in time to peak shortening (i.e. ‘primary mechanical dyssynchrony’). Primary electrical dyssynchrony is typical for LBBB and primary mechanical dyssynchrony is typical for regional ischaemia. The electrical and non-electrical aetiologies of dyssynchrony need to be differentiated since delay in electrical conduction may be the only aetiology amenable for CRT. We believe that clear differentiation between aetiologies is essential for understanding and applying dyssynchrony indices.
We recently introduced a novel method to estimate timing of electrical activation and thereby identify electrical dyssynchrony.33 This method utilizes LV pressure-strain loops to measure onset of active myocardial force generation (AFG). In an experimental preparation with implanted myocardial electrodes, we could show that time differences in onset AFG tracked very well regional differences in onset electrical activation since excitation–contraction coupling time was essentially constant. This was confirmed during different haemodynamic conditions, myocardial ischaemia, and LBBB. Figure 3 shows how to measure onset AFG. So far, the AFG method has only been used for research purposes and further testing is needed before it can be recommended for clinical use. One disadvantage is that invasive pressure is needed. Patients who are candidates for CRT, however, often need left heart catheterization and this provides access to LV pressure.
(Left panel) Definition of onset active force generation: recordings from a dog model in which onset active force generation was defined as the time of upward shift from segmental passive-elastic curve. (Right panels) Onset active force generation in the left ventricular lateral wall and septum in a patient with left bundle branch block. Onset of active force generation defined as the time that corresponded to the first marked upward deviation of the pressure-strain loop that resulted in a continued upward shift (circle). AFG = Active force generation.
We have recently introduced a novel, non-invasive method to measure LV pressure-strain loops which might have a potential to identify electrical dyssynchrony. The method is described in more detail in the last section of this article. As illustrated in Figure 4, regional LV pressure-strain loops are entirely different for segments with electrical dyssynchrony when compared with segments with reduced contractility due to ischaemia, although loop area and therefore segmental work is reduced in both cases. The most characteristic feature of electrical dyssynchrony is that the early-activated septum has marked pre-ejection shortening accompanied by pre-ejection lengthening in the lateral wall. An ischaemic segment, however, has typically pre-ejection lengthening and there is no associated pre-ejection lengthening in the opposite wall. It remains to be explored if pressure-strain loop configurations may be used this way to identify electrical dyssynchrony.
(A) Left ventricular pressure-strain loops during left bundle branch block from a dog experiment measured in the LV lateral wall and in the septum. In the lateral segment, the pressure-strain loop rotates counterclockwise, which is the normal pattern, and the area of the loop reflects segmental work. In the septal segment, loop area is markedly reduced relative to the lateral segment, which implies that septal work is markedly reduced. Part of the septal pressure-strain loop rotates clockwise, which means that work is performed on the septum by other segments. The septal loop also illustrates that pre-ejection septal shortening (red arrow) is active since it occurs against rising left ventricular pressure. (B) Left ventricular pressure-strain loops during left anterior descending coronary artery occlusion. The anterior segment is stretched in systole and the loop area is almost zero, indicating that the segment is affected by ischaemia and is not generating work. The black circle indicates the onset of active myocardial force generation.
Correct timing of septal activation
In patients with electrical dyssynchrony, the septum is usually activated first and timing of LV free wall activation is made with reference to the septum. It is therefore critical that timing of septal activation is done properly. Due to complex septal deformation patterns in LBBB, it is challenging to define the mechanical event which best reflects timing of septal activation. In normal hearts, the septum, similar to the LV free wall, shortens continuously during the LV ejection phase. During LBBB, however, septal deformation has two phases of systolic shortening, each followed by systolic lengthening as illustrated in Figures 5 and 6.34 Furthermore, as shown by Leenders et al.,35 this strain pattern may be modified by changes in contractility in the septum or LV free wall.
Systolic strain traces from the septum (thick lines) and lateral wall (thin lines) in a patient with left bundle branch block before (left) and after treatment with cardiac resynchronization therapy (right). During left bundle branch block without cardiac resynchronization therapy, there is a marked septal shortening during pre-ejection and there is shortening during mid-ejection (red). Each shortening is followed by lengthening, resulting in both early- and late-systolic lengthening (blue). The lateral wall shows pre-ejection lengthening before treatment (green). These changes are reverted after treatment. A pre-ejection stretch has appeared in the septum after treatment with cardiac resynchronization therapy, indicating that there is slight dyssynchrony also during cardiac resynchronization therapy. AVC = aortic valve closure; AVO = aortic valve opening; MVC = mitral valve closure; MVO = mitral valve opening. Modified from De Boeck et al.34
The study of Gjesdal et al.36 suggests that onset of pre-ejection septal shortening rather than onset of ejection shortening should be used for timing of septal activation. This is supported by the observation that onset R in local EMG coincides with the onset of septal shortening (Figure 6). Furthermore, the observation that pre-ejection septal shortening occurs against rising LV pressure is consistent with active septal contraction and implies that the septum is electrically activated during the pre-ejection phase (Figures 3, patient, and 4, experimental).
Pressures, septal segment length by sonomicrometry, septal intramyocardial electromyogram (EMG), and electrocardiogram (ECG) in a dog with left bundle branch block. The first vertical line indicates timing of electrical activation of the septum as measured by EMG. The next vertical lines indicate timing of aortic valve opening (AVO) and closure (AVC). There are two phases of septal systolic shortening and relengthening. Please note that onset septal shortening during ejection does not correspond to timing of electrical activation of the septum as measured by EMG. ED = end diastole.
Myocardial ejection velocities in assessment of dyssynchrony
Clinical studies which have investigated the ability of tissue Doppler echocardiography to identify responders to CRT have often measured peak myocardial velocity during LV ejection. This was also done in the PROSPECT-trial33 which concluded that no echocardiographic measure of dyssynchrony may be recommended to improve patient selection for CRT. As discussed in the previous paragraph of this article and as suggested by the studies of Russell et al.33 and Gjesdal et al.,36 ejection-phase velocities may be suboptimal measures of electrical dyssynchrony. There is no doubt that differences in timing of peak myocardial shortening velocity reflect mechanical dyssynchrony, but there are little data to support a strong relationship to electrical dyssynchrony. Figure 7, which is from an experimental study, illustrates how differences in timing of peak ejection velocity may lead to wrong conclusions regarding electrical dyssynchrony. It shows that peak septal ejection velocity occurs after peak lateral wall velocity during LBBB, whereas EMG by myocardial electrodes confirms that the septum is activated first. As they are yet not sufficiently proven to work, we suggest that ejection-phase velocities are not applied as standalone markers of electrical dyssynchrony in relation to CRT.
Representative traces from an experimental study showing velocity by sonomicrometry and regional electrical activation by intramyocardial electromyograms (EMG) for a septal and lateral LV wall segment during left bundle branch block. Note that peak septal ejection velocity occurs after peak lateral wall velocity, opposite to electrical activation by intramyocardial EMGs. AVC = aortic valve closure; AVO = aortic valve opening; IVC = isovolumic contraction.
Peak myocardial strain as a marker of dyssynchrony
As discussed above, recent studies indicate that myocardial strain may be superior to velocities for evaluation of dyssynchrony. Due to complex mechanical interactions between different parts of the LV wall, strain traces also need to be interpreted with caution. This limitation is illustrated by the two clinical examples in Figure 8. In Patient 1, different timing of peak strain in septum and LV lateral wall suggests marked dyssynchrony, whereas Patient 2 apparently has no significant dyssynchrony. Both patients have LBBB with QRS width >150 ms and no known coronary artery disease. In these circumstances, we would argue that it is highly unlikely for the septum to be activated after the lateral wall as suggested by timing of peak strain for Patient 2. This probably reflects dependency of timing of peak strain on other factors than electrical conduction.35
Strain traces by speckle-tracking echocardiography in four-chamber views showing peak septal strain and peak LV lateral wall strain (arrows) during left bundle branch block for a patient with peak septal strain during isovolumic contraction (Patient 1) and a patient with peak septal strain during late ejection (Patient 2). Note that for Patient 2, peak septal strain occurs after peak lateral wall strain. Both patients have QRS width >150 ms and no known or prior history of coronary artery disease.
Expeimental studies suggest that strain derived dyssynchrony indices may exaggerate underlying electrical dyssynchrony. Prinzen et al..38 have shown that delay in onset shortening between early and late activated LV segments exceeds the delay in electrical conduction during pacing-induced dyssynchrony. In a recent study,38 we showed that the apparent electromechanical delay in the LV lateral wall during LBBB is due to higher rate of LV pressure rise (due to early septal contraction) at the time of lateral wall activation, resulting in a longer time before the segment generates active force at a rate superior to the load rise which is a requirement for shortening. These findings suggest that LV contractility may modify timing of onset of shortening and hence mechanical dyssynchrony.
Assessing regional work: a new approach for understanding dyssynchrony
LV dyssynchrony is associated with marked inter-segment differences in myocardial work37 as illustrated in Figures 3, right panel, and 4. During LBBB, the septum performs little and in some cases negative work, whereas the LV lateral wall is hyperactive due to pre-systolic stretching (increased preload) caused by the early-activated septum (Figure 4). Analysis of regional work during electrical dyssynchrony may provide important insights into mechanisms of remodelling and LV dysfunction.37,39,40 Abnormal inter-segmental work distribution during LBBB is associated with differences in regional blood flow and oxygen demand and may also account for remodelling of the LV with hypertrophy of the hyperactive LV lateral wall and thinning og the early-activated septum.40–43
We have recently introduced a non-invasive method to assess regional myocardial work by LV pressure-strain loops analysis.44 A non-invasive LV pressure is obtained by using an empiric, normalized reference curve which is adjusted according to duration of the isovolumic and ejection phases determined by echocardiography, and absolute pressure level is estimated from brachial artery cuff pressure. The non-invasive LV pressure-strain loop area gives an accurate quantification of regional work when using invasive LV pressure-strain loops as the reference method,44 and the method is currently undergoing clinical testing. Assessment of regional work may provide important insights into cardiac mechanics and may be useful when evaluating patients who are candidates for CRT.
There is no doubt that echocardiography represents a unique tool for quantification of LV dyssynchrony, but further studies are needed before a conclusion can be reached regarding its role in selection of patients for CRT. We recommend that new developments focus on methods that can differentiate between electrical and non-electrical aetiologies of dyssynchrony and on methods to quantify its negative mechanical impact. Ultimately, only patients with an electrical aetiology are likely to respond favourably to CRT. Just measuring velocity indices does not identify the aetiology. Myocardial strain appears more promising, but one should be aware that timing of peak systolic strain is determined not only by electrical conduction. One should also take into account potential ischaemia, scarring, and other structural changes as contributors to dyssynchrony. The proposed AFG method to identify electrical dyssynchrony and LV pressure-strain loop analysis to quantify the impact of dyssynchrony on regional work may be worth exploring in future studies. Furthermore, it is important to be aware that LV dyssynchrony may have a combination of aetiologies, not all amenable for CRT.
Conflict of interest: none declared.
Dr Russell was recipient of a clinical research fellowship from the University of Oslo.
. 2010 Focused Update of ESC Guidelines on device therapy in heart failure: an update of the 2008 ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure and the 2007 ESC Guidelines for cardiac and resynchronization therapy: developed with the special contribution of the Heart Failure Association and the European Heart Rhythm Association. Europace 2010;12:1526-36.
. 2009 focused update: ACCF/AHA guidelines for the diagnosis and management of heart failure in adults: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines: developed in collaboration with the International Society for Heart and Lung Transplantation. Circulation 2009;119:1977-2016.
. 2009 Focused update incorporated into the ACC/AHA 2005 guidelines for the diagnosis and management of heart failure in adults a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines: developed in collaboration with the International Society for Heart and Lung Transplantation. J Am Coll Cardiol 2009;53:e1-90.
. Echocardiography for cardiac resynchronization therapy: recommendations for performance and reporting—a report from the American Society of Echocardiography Dyssynchrony Writing Group endorsed by the Heart Rhythm Society. J Am Soc Echocardiogr 2008;21:191-213.
. Clinical efficacy of cardiac resynchronization therapy using left ventricular pacing in heart failure patients stratified by severity of ventricular conduction delay. J Am Coll Cardiol 2003;42:2109-16.
. Cardiac resynchronization therapy in asymptomatic or mildly symptomatic heart failure patients in relation to etiology: results from the REVERSE (REsynchronization reVErses Remodeling in Systolic Left vEntricular Dysfunction) study. J Am Coll Cardiol 2010;56:1826-31.
. Tissue Doppler imaging is superior to strain rate imaging and postsystolic shortening on the prediction of reverse remodeling in both ischemic and nonischemic heart failure after cardiac resynchronization therapy. Circulation 2004;110:66-73.
. Tissue Doppler echocardiographic evidence of reverse remodeling and improved synchronicity by simultaneously delaying regional contraction after biventricular pacing therapy in heart failure. Circulation 2002;105:438-45.
. Relationship between endocardial activation sequences defined by high-density mapping to early septal contraction (septal flash) in patients with left bundle branch block undergoing cardiac resynchronization therapy. Europace 2012;14:99-106.
. Assessment of left ventricular dyssynchrony by speckle tracking strain imaging comparison between longitudinal, circumferential, and radial strain in cardiac resynchronization therapy. J Am Coll Cardiol 2008;51:1944-52.
. Guidelines for cardiac pacing and cardiac resynchronization therapy: the Task Force for Cardiac Pacing and Cardiac Resynchronization Therapy of the European Society of Cardiology. Developed in collaboration with the European Heart Rhythm Association. Eur Heart J 2007;28:2256-95.
. Evaluation of left ventricular dyssynchrony by onset of active myocardial force generation: a novel method which differentiates between electrical and mechanical etiologies. Circ Cardiovasc Imaging 2010.
. Evaluation of left ventricular dyssynchrony by onset of active myocardial force generation: a novel method that differentiates between electrical and mechanical etiologies. Circ Cardiovasc Imaging 2010;3:405-14.
. Asymmetric thickness of the left ventricular wall resulting from asynchronous electric activation: a study in dogs with ventricular pacing and in patients with left bundle branch block. Am Heart J 1995;130:1045-53.