Sequential Versus Simultaneous Biventricular Resynchronization for Severe Heart Failure
Evaluation by Tissue Doppler Imaging
Background— Cardiac resynchronization therapy (CRT) by means of simultaneous biventricular pacing improves left ventricular systolic performance and synchrony in patients with heart failure and bundle-branch block. We used tissue tracking and 3D echocardiography to evaluate the impact of sequential CRT with individualized interventricular delay programming.
Methods and Results— Twenty consecutive patients with severe heart failure and left bundle-branch block were included. Tissue tracking and 3D echocardiography were carried out before and on the day after pacemaker implantation. Eleven different interventricular delays were examined in each patient. Patients were reexamined after 3 months. Simultaneous CRT immediately reduced the extent of myocardium displaying delayed longitudinal contraction (DLC) from 48.6±16% to 23.2±13% (P<0.01) and increased left ventricular ejection fraction percentage (LVEF%) from 22.4±6% to 29.7±5% (P<0.01). However, optimum sequential CRT caused a further reduction in the extent of DLC from 23.2±13% to 11.1±7.2% (P<0.01), with a simultaneous increase in LVEF% (from 29.7±5% to 33.9±6%, P<0.01). Three months of optimum sequential CRT further improved LVEF% (from 33.6±6% to 38.6±7.2%, P<0.01). Tissue tracking detected the segments with DLC, and their location determined optimum interventricular delay programming. Compared with simultaneous CRT, sequential CRT increased diastolic filling time by 7±2.5%.
Conclusions— Compared with simultaneous CRT, sequential CRT significantly improves left ventricular systolic and diastolic performance. Tissue tracking can be used to select optimum interventricular delay during CRT.
Received June 12, 2002; revision received August 7, 2002; accepted August 7, 2002.
Cardiac resynchronization therapy (CRT) with simultaneous right ventricular (RV) and left ventricular (LV) stimulation is a promising therapeutic option in patients with severe heart failure (HF) and left bundle-branch block (LBBB).1 We and others have previously reported on the use of tissue Doppler imaging with tissue tracking (TT) for the detection and quantification of LV systolic performance2–4⇓⇓ and LV asynchrony.3,4⇓ We found that the extent of myocardium with delayed longitudinal contraction (DLC) predicted the improvement in LV systolic performance and reversion of LV remodeling during short- and long-term CRT.3,4⇓ Our observations indicate that DLCs represent mechanical LV asynchrony and thus a contractile reserve, which can be recruited by means of CRT.4
However, in patients with HF and LBBB, the location of myocardium displaying DLCs may vary. Recent development in pacemaker technology enables separate activation of the ventricular leads. In the present study, we hypothesize that individually tailored preactivation of myocardium displaying DLCs would further improve the overall response to CRT.
Twenty-one consecutive patients referred for implantation of a biventricular pacemaker were included in the present study after oral and written informed consent. The study was approved by the local ethics committee and was conducted according to the Helsinki declaration.
All patients were in sinus rhythm and had LBBB with a QRS duration >130 ms. Patients were in New York Heart Association (NYHA) class III or IV despite contemporary medical treatment for HF (Table).
Three transvenous pacing leads were inserted, one in the right atrium and another on the high interventricular septum or in the RV outflow tract. In addition, a coronary sinus lead (Medtronic 2187 or 10512) was positioned on the LV free wall through a coronary sinus tributary. The RV and LV pacing leads were positioned to ensure that the sensed local intracardiac activation was measured early for the RV lead and late for the LV lead according to the QRS complex on the surface ECG. The pacing leads were connected to a dual-chamber biventricular pacemaker (In-Sync III, Medtronic) with programmable interventricular delay. The pacemaker was programmed in the DDD mode, and adjustment of the atrioventricular (AV) delay was performed during simultaneous pacing. According to Ritter et al,5 the longest possible AV filling time, without truncation of the A wave, was chosen by means of pulsed Doppler analysis of the transmitral flow. In these patients, the average AV delay was 106.5±26.6 ms.
We have previously reported in detail on the tissue Doppler imaging techniques and 3D echocardiography (ECHO) in the setting of biventricular pacing as well as on observer variability.3,4,6–8⇓⇓⇓⇓ Intraobserver variability for measurement of LV volumes and LV ejection fraction percentage (LVEF%) is 3% to 5%. 3D ECHO volumes are accurate and reproducible compared with magnetic resonance imaging6; consequently, the number of patients required in a 2D ECHO study should be at least 4 times higher to achieve the same statistical power on volume measurements.7 For myocardial shortening assessed by TT, the intraobserver variability is 6%.
In brief, 3D ECHO and tissue Doppler recordings were performed on the day before and after implantation of the pacemaker. Measurements were repeated after 3 months. During the first postimplantation evaluation, the ECHO examinations were carried out during simultaneous CRT and at 5 different interventricular delay intervals (12, 20, 40, 60, and 80 ms) with either LV lead preactivation or RV lead preactivation. Thus, a total of 11 different interventricular delays were examined, with an equilibrium period of 10 minutes between each examination.
3D ECHO was performed during end-expiratory apnea within 1 breath-holding by using ECG-triggered coaxial rotation from the apical window with 30° intervals between the scanning planes. The resulting 6 digital cine loops were transferred to a computer for offline analysis (Echo-Pac software, GE-Vingmed Ultrasound). In each of the 6 views, endocardial borders were drawn manually in end diastole and end systole for calculation of LV end-diastolic volume (LVEDV), end-systolic volume (LVESV), and LVEF%.3,4,6–8⇓⇓⇓⇓
Tissue Doppler recordings were acquired as digital loops in end-expiratory apnea during 1 heart beat in the apical 4-chamber, 2-chamber, and long-axis views. To avoid aliasing, the settings of the ultrasound equipment and color-coded area were adjusted to obtain the highest possible frame rate (mean 115 frames/s, range 100 to 135 frames/s). A TT image in systole was derived from the digital loops in each of the apical views (Figures 1A and 2⇓A, top). Another TT image was recorded in diastole to visualize myocardium with DLC (Figures 1A and 2⇓A, bottom), and the number of segments with DLC was registered. Strain rate analysis (Figure 3) was carried out in each segment, and motion toward the apex in diastole was registered as DLC if negative strain rate analysis documented that the motion reflected true shortening.
The 16-segment LV model9 was applied to the TT images. From the color-coded TT image, the motion amplitude toward the apex in systole was recorded in each segment. The global systolic contraction amplitude (GSCA) was calculated as the average shortening amplitude of all 16 segments. Calculation of GSCA was performed in the 11 different interventricular delay recordings at baseline and after 3 months. Optimum interventricular delay programming of the pacemaker was identified by maximum GSCA in each patient (Figures 1B and 2⇑B).
To minimize the variability of the measurements, all ECHO recordings were performed and analyzed by the same author (P.S.). A 6-minute hall walk was performed at baseline and after 3 months.
Differences between ECHO variables at baseline and during CRT were evaluated by a nonparametric ANOVA (Kruskal-Wallis). A paired t test was used to evaluate changes in the 6-minute hall walk and NYHA class. A value of P<0.05 was considered statistically significant.
One patient developed atrial fibrillation between pacemaker implantation and ECHO evaluation and was excluded from analysis. Clinical and demographic variables of the remaining 20 patients are given in the Table. Tissue Doppler analysis was feasible in all 16 segments in all patients. Similarly, 3D ECHO recordings were adequate for analysis in all 20 patients.
Location of DLC
The present study population included 11 patients with ischemic cardiomyopathy caused by previous myocardial infarction and 9 patients with idiopathic dilated cardiomyopathy. In spite of comparable LBBB configuration, the locations of DLCs differed between the 2 groups of patients. In patients with idiopathic dilated cardiomyopathy, myocardium with DLC tended to be located in the lateral and posterior walls of the LV (Figure 1A, bottom). In contrast, DLC was more frequent in the septum and in the inferior wall in patients with ischemic cardiomyopathy (Figure 2A, bottom). However, 1 patient with idiopathic dilated cardiomyopathy showed DLC in the septum and the adjacent part of the inferior wall, and 1 patient with ischemic cardiomyopathy presented with DLC in the lateral and posterior walls (Figure 4A and 4B).
Baseline Versus Simultaneous CRT
Simultaneous CRT reduced the extent of segments at the LV base, displaying DLCs from 48.6±16% to 23.2±13% (P<0.01). In parallel, the GSCA index increased by 32.6±6% (P<0.01). This was accompanied by an increase in LVEF% from 22.4±6% to 29.7±5% (P<0.01) (Figure 5), and LVEDV and LVESV were reduced by 9±5% and 18±7%, respectively (P<0.01 for both) (Figure 5).
The diastolic filling time as measured by pulsed Doppler evaluation increased from 380±80 ms at baseline to 430±88 ms during simultaneous CRT with optimized AV delay (P<0.01).
Simultaneous Versus Optimized Sequential CRT
In all patients, the ECHO parameters improved further during sequential CRT. Preactivation of the LV lead was superior in 9 patients (exemplified in Figure 1B), whereas preactivation in the RV was superior in the remaining 11 patients (exemplified in Figure 2B). Compared with simultaneous CRT, optimum sequential CRT reduced the extent of segments with DLC at the LV base from 23.2±13% to 11.1±7.2% (P<0.05). GSCA was increased by 18±6% (P<0.01), accompanied by an increase in LVEF% from 29.7±5% to 33.6±6% (P<0.01) (Figure 5). The LVEDV remained unchanged at optimum sequential CRT, whereas LVESV was further reduced (9±4%) (P<0.05), as seen in Figure 5.
Without any further AV-delay optimization, the diastolic filling time increased from 430±88 ms during simultaneous CRT to 460±80 ms (P<0.05) during optimum sequential CRT.
Predictors of Optimum Sequential CRT
In the 9 patients with DLC of the posterolateral wall, optimum sequential CRT was obtained through LV lead preactivation. In contrast, RV lead preactivation resulted in the highest GSCA in the 11 patients with DLC in the septum and adjacent inferior wall.
Average GSCAs from all 11 different interventricular delays are presented in Figure 6, dividing patients into those showing optimum response from RV preactivation versus those showing optimum response from LV preactivation. Optimum RV and LV preactivation intervals were narrow, ranging between 12 and 20 ms. An increase in interventricular delay to ≥40 ms caused a significant reduction in GSCA (P<0.01), even though the lead to preactivate was chosen correctly. This timing of the interventricular delay also resulted in a poorer GSCA than obtained during simultaneous pacing (P<0.01). Any preactivation of the lead opposite the one showing optimum resynchronization caused a significant reduction in GSCA compared with simultaneous CRT (P<0.01).
All of the 20 patients were alive after 3 months. The NYHA class improved from 3.45±0.5 at baseline to 1.9±0.45 after 3 months. Similarly, the 6-minute hall walk improved from 222±100 m at baseline to 401±110 m after 3 months (P<0.01). During 3 months of follow-up, LVEDV and LVESV were significantly reduced (from 218±58 to 194±59 mL and from 145±47 mL to 119±56 mL, respectively; P<0.01 for both). LVEF% increased from 33.6±6% to 38.6±7% (P<0.01), as did GSCA (from 4.7±0.5 to 5.5±0.6 mm, P<0.01).
The present study confirms our previous observations of an immediate beneficial effect of simultaneous CRT on LV volumes and systolic performance.3,4,8⇓⇓ It is also in accordance with our earlier findings that this improvement appears in parallel with a reduction in the extent of myocardium displaying DLC with a negative strain rate, ie, segments showing contraction in diastole.3,4⇓
Despite similar QRS morphology, patients with HF and LBBB may present with a different location of mechanical asynchrony, which primarily seems to be related to the underlying etiology (Figure 4A and 4B). In the present study, individual tailoring of the interventricular delay with preactivation of regions showing mechanical asynchrony produced a further significant reduction in the extent of DLC, ie, improved recruitment of the contractile reserve. This was accompanied by a significant improvement in LVEF%. Thus, compared with simultaneous CRT, sequential CRT offers a potential to further improve LV systolic performance, and this additional benefit can easily be visualized and quantified by means of TT. The improvement in GSCA, as demonstrated by TT, is closely related to the improvement in LVEF as previously reported.4
Surprisingly, the benefit of sequential CRT was present within a relatively short range of interventricular delays from RV preactivation by 20 ms to LV preactivation by 20 ms. Any further increase in interventricular delay or preactivation of regions not exhibiting DLCs resulted in increased mechanical asynchrony and compromised systolic performance, even when sequential CRT was compared with simultaneous CRT (Figure 6). These observations emphasize the importance of a preimplantation evaluation of the mechanical asynchrony in terms of the location of myocardium with DLCs. For that purpose, TT is an excellent “at a glance method” that is also very useful for customizing the pacemaker programming (Figures 1 and 2⇑).
During short-term pacing, a further improvement in LV systolic performance and dimensions was noted. This is in accordance with our previous findings of reversed LV remodeling during 1 year of CRT4 and in accordance with another short-term evaluation study.10 In parallel, NYHA class improved significantly, and the 6-minute hall walk increased to a doubled walking distance, an improvement that exceeds recent reports.1,10⇓
AV delay adjustment is commonly performed by using pulsed Doppler evaluation of transmitral flow,1,10⇓ as described by Ritter et al.5 Previous studies have emphasized the importance of a short AV delay during standard dual-chamber pacing in patients with HF to optimize the hemodynamic response.11,12⇓ However, in another study, the same benefit could not be documented.13 At present, the importance of AV-delay programming during CRT has systematically been evaluated by dP/dt measurements in 1 study only.14
Obviously, DLC not only diminishes the global LV systolic performance but also interferes with diastolic filling and causes abnormal LV wall stress in diastole. In the present study, establishment of simultaneous CRT together with AV-delay adjustment significantly prolonged the diastolic filling time. Sequential CRT with individually optimized interventricular delay caused an additional significant increase in diastolic filling time. This benefit appeared in parallel with a further reduction in the extent of segments displaying DLC, without any further adjustment of the AV delay or change in heart rate. Thus, the degree of diastolic filling abnormality seemed related to the degree of LV asynchrony. This indicates that proper correction of LV asynchrony can improve AV mechanics and diastolic LV performance.15 Thus, optimized ventricular synchronization should probably be achieved in advance of AV-delay adjustment to prevent inappropriate AV-delay programming, which might reduce LV filling.15
In conclusion, in patients with severe HF and LBBB, tailored sequential CRT, compared with simultaneous CRT, causes a further significant improvement in LV systolic and diastolic performance. Tissue Doppler imaging in terms of TT proved useful in predicting and optimizing interventricular delay programming during sequential biventricular pacing.
- ↵Pan C, Hoffmann R, Kuhl H, et al. Tissue tracking allows rapid and accurate visual evaluation of left ventricular function. Eur J Echocardiogr. 2001; 2: 197–202.
- ↵Ritter P, Padeletti L, Gillio-Meina L, et al. Determination of the optimal atrioventricular delay in DDD pacing: comparison between echo and peak endocardial acceleration measurements. Europace. 1999; 1: 126–130.
- ↵Kim WY, Sogaard P, Mortensen PT, et al. Three-dimensional echocardiography documents haemodynamic improvement by biventricular pacing in patients with severe heart failure. Heart. 2001; 85: 514–520.
- ↵Schiller NB, Shah PM, Crawford M, et al. Recommendations for quantitation of the left ventricle by two-dimensional echocardiography: American Society of Echocardiography Committee on Standards, Subcommittee on Quantitation of Two-Dimensional Echocardiograms. J Am Soc Echocardiogr. 1989; 2: 358–363.
- ↵Yu CM, Chau E, Sanderson JE, et al. 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: 438–445.
- ↵Auricchio A, Sommariva L, Salo RW, et al. Improvement of cardiac function in patients with severe congestive heart failure and coronary artery disease by dual chamber pacing with shortened AV delay. Pacing Clin Electrophysiol. 1994; 17(pt 1): 995–997.
- ↵Auricchio A, Stellbrink C, Block M, et al. Effect of pacing chamber and atrio-ventricular delay on acute systolic function of paced patients with congestive heart failure: the Pacing Therapies for Congestive Heart Failure Study Group: the Guidant Congestive Heart Failure Research Group. Circulation. 1999; 99: 2993–3001.