17 Echocardiography and Its Role in Cardiac Resynchronization Therapy
Cardiac resynchronization therapy (CRT) has an established role in the care of people with severe left ventricular (LV) systolic dysfunction, heart failure (HF) symptoms, and evidence of electrical dyssynchrony manifest as a wide QRS complex on the 12-lead electrocardiogram (ECG).1–7 The main principle of CRT is to attempt to normalize the electrical and mechanical ventricular activation sequence by strategically positioning a pacing lead in a coronary sinus tributary. This positioning allows for LV pre-excitation and may favorably alter the timing of LV events, thereby improving LV systolic function. Currently established indications for CRT include LV ejection fraction (LVEF) <35%, New York Heart Association (NYHA) class III or IV HF symptoms, and ECG evidence of dyssynchrony with a QRS duration of >120 ms.8 People selected for CRT in clinical trials using these criteria have shown significant improvement in NYHA class, 6-minute walk distance, peak oxygen consumption (VO2 max), reduced HF hospitalizations, and reduced total mortality.1–7 In addition, people undergoing CRT have shown significant improvement in LV reverse-remodeling parameters, such as a reduction in LV end-systolic volume and mitral regurgitation (MR).9 Unfortunately, not all patients with HF benefit from CRT. Overall, approximately 25% to 35% of patients selected for CRT using standard criteria do not show objective improvement over long-term follow-up.10 Since the advent of CRT echocardiography techniques have evolved in an attempt to further refine the selection, optimization, and follow-up of CRT candidates. This chapter summarizes the current role of echocardiography in patients referred for CRT.
Precardiac Resynchronization Therapy Patient Selection Using Echocardiography
The main goal of echocardiography in the pre-CRT implantation time frame is to identify and quantify dyssynchrony. Dyssynchrony refers to an alteration in both the electrical and mechanical cardiac activation sequences. Three types of dyssynchrony pertain to echocardiography and CRT: inter-, intra-, and atrioventricular.11 Interventricular dyssynchrony refers to an alteration in the timing sequence between the right and left ventricles. Interventricular dyssynchrony may be identified using conventional echo-Doppler methods preimplantation and does have modest predictive value in terms of predicting CRT response.7 Intraventricular dyssynchrony refers to alteration of the activation sequence within the left ventricular myocardium. For example, in left bundle branch block (LBBB) the interventricular septum is often activated significantly before the LV posterior and lateral walls. Such dyssynchrony may lead to impaired coordination of myocardial contractility, decrease in LVEF, and MR.12 Quantification of pre-CRT intraventricular dyssychrony may assist with pre-CRT selection of patients. Atrioventricular dyssynchrony refers to an alteration in the timing of atrial and ventricular activation and contraction that may have adverse effects on cardiac function. For example, first-degree atrioventricular (AV) block may lead to a significant delay between the end of atrial contraction and the onset of ventricular contraction. In HF patients, such a delay can result in higher left ventricular end-diastolic pressure (LVEDP) and diastolic MR.
Quantification of Interventricular Dyssynchrony
Measurement of interventricular dyssynchrony aims to quantify the temporal difference between the onset of right ventricle (RV) and LV activation by measurement of both LV and RV pre-ejection intervals. Some techniques report only the left ventricular pre-ejection interval (LVPEI),7 whereas others report the difference between the right ventricular pre-ejection interval (RVPEI) and the LVPEI to yield a measurement of interventricular dyssynchrony.11
Technique
Using conventional Doppler imaging, pulsed-wave Doppler measurements are obtained sequentially in the right ventricular outflow tract (parasternal short-axis view) and left ventricular outflow tract (apical five-chamber view). Continuous ECG measurement is recorded. For each outflow tract, the time from the onset of the QRS complex to the onset of Doppler flow is measured (Fig. 17-1). The difference between the measurements is recorded as the measurement of interventricular dyssynchrony. A difference of >40 msec is considered evidence of significant interventricular dyssynchrony.11 LVPEI measured >140 msec is also considered evidence of significant interventricular dyssynchrony.7,10
Evidence
In the Cardiac Resynchronization in Heart Failure (CARE-HF) study,7 quantification of interventricular dyssynchrony was used as an inclusion criterion for study participants with a QRS width from 120 to 149 msec. In a subgroup analysis of this study, the presence of interventricular dyssynchrony with a cut-off value of ≥49.2 msec in those subjects who received CRT predicted a lower hazard ratio (HR) (0.50; 0.36–0.70) of outcome events when compared to those without interventricular dyssynchrony (0.77; 0.58–1.02). However, in terms of predicting CRT response, results have been variable. The Predictors of Response to CRT (PROSPECT) study13 found that in a multicenter setting, the predictive value of interventricular dyssynchrony was modest. An interventricular mechanical dyssynchrony of >40 msec was only 55.2% sensitive and 56.4% specific for CRT response, yielding an area-under-the-curve (AUC) of 0.58. Similarly, measurement of LVPEI was only 66.3% sensitive and 47.1% specific, yielding an AUC of 0.60. The PROSPECT investigators suggested that measurements of interventricular dyssynchrony should not be used as a basis for clinical decisions regarding CRT implantation.
Quantification of Intraventricular Dyssynchrony
Intraventricular dyssynchrony is present in a significant majority of people with HF and wide QRS complexes. The proportion of individuals with dyssynchrony increases as QRS duration increases and is especially prevalent when the QRS duration is >150 msec. The primary aim of a pre-CRT echocardiogram for intraventricular dyssynchrony is to identify those individuals who do not have significant mechanical dyssynchrony and therefore may be less likely to respond to CRT. Furthermore, echocardiography may identify a subset of individuals with QRS duration <120 msec who have significant mechanical dyssynchrony and therefore may benefit from CRT. However, the Resynchronization Therapy In Normal QRS (RETHINQ) trial,14 published in 2007, failed to demonstrate a significant benefit of CRT in people with NYHA class III HF symptoms, severe LV systolic dysfunction, QRS duration <120 msec, or echocardiographic evidence of mechanical dyssynchrony. There are registry and other nonrandomized data that suggest people with HF and wide QRS but no evidence of dyssynchrony have a relatively poor prognosis.11 In addition, in the absence of identifiable dyssynchrony, they may be less likely to respond to CRT. However, the PROSPECT study, published in 2008, suggested that the sensitivity, specificity, and AUC of current echocardiographic techniques to identify CRT nonresponders are inadequate to influence clinical decision-making.13 Other than the absence of dyssynchrony, there are several potential reasons for nonresponse to CRT. Extensive myocardial scar, suboptimal LV lead placement, and post-CRT myocardial infarction may influence CRT response.13 Newer techniques that combine both radial and longitudinal evaluation of dyssynchrony have improved sensitivity and specificity to identify dyssynchrony or its absence.15 Radial speckle tracking16 and real-time three-dimensional echocardiography (RT3DE)17 show some promise as improved techniques for precisely quantifying intraventricular dyssynchrony.
Two-Dimensional Imaging Technique
In some people with intraventricular dyssynchrony, certain hallmark features are seen on the two-dimensional (2D) images. Early septal contraction, often described as a septal flash or bounce, has been described. However, this method lacks sensitivity; therefore, more quantitative methods are required.10
M-Mode Imaging Techniques
M-mode imaging is useful for temporal resolution of the synchrony of LV wall motion. Pitzalis et al.18 have described a technique that quantifies the difference in timing of contraction between the septal and posterior walls. Using this technique, septal-to-posterior wall motion delay (SPWMD) of >130 msec defines intraventricular dyssynchrony (Fig. 17-2).
Technique
2D images are obtained of the septal and posterior walls in the parasternal long-axis view. M-mode imaging is activated. Continuous ECG recording is obtained. Using the M-mode images, the time difference between the onset of septal wall contraction and the onset of posterior wall contraction is obtained. Some investigators also perform this M-mode measurement using the parasternal short-axis 2D image. A similar analysis of the M-mode images is performed. Color M-mode may, in some cases, improve temporal resolution.
Evidence
The evidence for this technique is weak. Two studies done at the same center18,19 showed an apparently significant predictive value of the SPWMD to predict CRT response as measured by reverse remodeling and symptomatic improvement. However, a study published in 2005 that evaluated the predictive value of the SPWMD in a large cohort from the VENTAK CHF/CONTAK CD Biventricular Pacing Study did not find this measurement to be predictive of CRT response.20 In the PROSPECT study, only 60.7% of echocardiograms were evaluable for SPWMD.13 The sensitivity and specificity for predicting CRT response were 6.3% and 91.7% for clinical improvement and 9.5% and 92.9% for improvement in left ventricular end-systolic volume (LVESV), yielding an AUC of 0.52 and 0.50, respectively.13 This technique is significantly limited by the fact that nearly 40% of patients in a multicenter setting were not evaluable. In people with nonischemic cardiomyopathy, the technique may have more utility. However, the predictive value overall for CRT response using this technique is poor.
Tissue Doppler Imaging Techniques
Tissue Velocity Imaging
Tissue Doppler (TD) imaging techniques measure the velocity of myocardial contraction at the base and mid-regions of the LV in the longitudinal plane. These measurements are performed in the standard apical four-chamber, two-chamber, and three-chamber long-axis views. There are two types of tissue velocity imaging (TVI): color-coded and pulsed. The preferred method is color-coded TVI because it allows for offline analysis of data, and this is the technique that is reviewed here. A 1-cm sampling volume is obtained in each region of interest, and the tissue velocity in this region is sampled. In this way, differences in tissue velocity between opposing walls in each view can be obtained (Fig. 17-3).15 Alternatively, all longitudinal velocity measurements may be compared by calculating the standard deviations among the 12 segments measured.21 These methods are used to quantify intraventricular dyssynchrony of the left ventricle. Using the Bax method, any difference in opposing ventricular walls >65 msec is defined as intraventricular dyssynchrony.15 The corresponding cut-off for intraventricular dyssynchrony using the Yu index is a standard deviation of 12 segments >33 msec.21
Color Doppler Tissue Analysis
Identify components of the time–velocity curve to ensure physiologic signal quality. These components include the isovolumic contraction velocity (<60 msec from QRS onset), the systolic (S) wave (a positive deflection occurring during LV ejection), the early diastolic (E) wave, and the late diastolic (A) wave, both negative deflections (Figs. 17-4 and 17-5)
Evidence
The simplest approach to TVI involves measuring the basal segments of the apical four-chamber view to measure the septal-to-lateral delay. Bax et al.15 established a four-segment model that involved the basal segments of the septal, lateral, inferior, and anterior LV walls.15 An opposing wall delay of ≥65 msec predicted clinical response to CRT (improvement in NYHA class and 6-minute walking distance) as well as reverse remodeling (≥15% reduction in LVESV). Similarly, subjects with LV dyssynchrony ≥65 msec had a better prognosis post-CRT when compared to subjects without dyssynchrony.15 Yu et al.21 developed a 12-segment model that derives information from all three apical LV views. The Yu index is derived by calculating the standard deviation (SD) of the time to peak systolic velocity in the LV ejection phase. The 12-site SD cutoff value to define dyssynchrony is 33 msec.21 This value was obtained using measurements from healthy subjects. To predict LV reverse remodeling (≥15% reduction in LVESV) in subjects with QRS duration >150 msec, the cutoff SD value of 33 msec has a sensitivity of 83% and a specificity of 86%.21 In the multicenter PROSPECT study, TVI did not perform as well in predicting CRT clinical and reverse remodeling outcomes. For the Bax method, 66.8% of echocardiograms were evaluable.13 The predictive values for clinical composite score improvement and for improvement in LVESV were poor, with an AUC of 0.50 and 0.61, respectively.13 The Yu method was difficult to reproduce in a multicenter setting, with only 50% of echocardiograms evaluable.13 AUC for improvement in clinical composite score was 0.60; for improvement in LVESV it was 0.55.13 Thus, it appears that performance of these methods in single centers where the technique has been practiced and refined is much superior to a multicenter setting in terms of prediction of CRT outcomes.
Tissue synchronization imaging (TSI) uses color-coded time-to-peak-velocity data and superimposes this information on the standard 2D apical LV images. In this way, a visual rendering of the latest activated myocardial region is displayed. It is important to focus on peak tissue velocity during the early ejection period and to exclude early isovolumic contraction and late post-systolic shortening. Gorscan et al.22 used color TSI to guide placement of regions of interest and assess an anteroseptal-to-posterior wall delay ≥65 msec. This technique was able to predict an acute improvement in stroke volume post-CRT. Yu et al. also used TSI in 56 patients and found the SD of 12-segment tissue velocities had a receiver operating curve of 0.90 to predict CRT response.23 However, as shown in the PROSPECT study, such techniques may not perform as well in a multicenter setting.
