Author + information
- Received March 27, 2002
- Revision received July 17, 2002
- Accepted August 1, 2002
- Published online December 4, 2002.
- Laurent Fauchier, MD, PhD*,*,
- Olivier Marie, MD*,
- Danielle Casset-Senon, MD†,* (, )
- Dominique Babuty, MD, PhD*,
- Pierre Cosnay, MD* and
- Jean Paul Fauchier, MD, FACC*
- ↵*Reprint requests and correspondence:
Dr. Laurent Fauchier, Service de Cardiologie B, Centre Hospitalier Universitaire Trousseau, 37044 Tours, France.
Objectives The study evaluated the prognostic value of interventricular and intraventricular dyssynchrony in idiopathic dilated cardiomyopathy (IDC).
Background Biventricular pacing is an emerging treatment for patients with dilated cardiomyopathy and ventricular dyssynchrony. The prognostic values of interventricular and intraventricular dyssynchrony have not been previously compared.
Methods A total of 103 patients with IDC were studied. Left bundle branch block was present in 25% of patients. Equilibrium radionuclide angiography was performed and Fourier phase analyses were examined in both ventricles. Difference between the mean phase of left ventricle (LV) and right ventricle (RV) assessed interventricular dyssynchrony, and standard deviations (SDs) of the mean phase in each ventricle assessed intraventricular dyssynchrony.
Results The QRS duration was related to both interventricular and intraventricular dyssynchrony. A degradation of the hemodynamic status was associated with an increase in intraventricular dyssynchrony but not in interventricular dyssynchrony. With a follow-up of 27 ± 23 months, 18 patients had a major cardiac event (7 cardiac deaths; 11 worsening, leading to heart transplantation). The SDs of the LV and RV mean phase and QRS duration were predictors of cardiac event (all p < 0.0001), but interventricular dyssynchrony was not. Among 13 univariate predictors of cardiac event, the only independent predictors were an increased SD of LV mean phase (p = 0.0004) and an increased pulmonary capillary wedge pressure (p = 0.009).
Conclusions Intraventricular dyssynchrony evaluated with phase analysis of radionuclide angiography is an independent predictor of cardiac event in IDC. The prognosis is related to intraventricular rather than to interventricular dyssynchrony in IDC.
Biventricular pacing (BiV) is an emerging treatment for patients with severe dilated cardiomyopathy and ventricular dyssynchrony. Recent studies have demonstrated improvements in hemodynamic status during atrial synchronized BiV pacing. In the randomized, controlled crossover Multisite Stimulation in Cardiomyopathy (MUSTIC) study in patients with severe heart failure (HF) due to chronic left ventricular systolic dysfunction, with QRS >150 ms, 6-min walk distance was 23% greater with three months of active BiV, whereas quality-of-life score improved and peak oxygen uptake increased by 8% (1). Phase analysis of radionuclide angioscintigraphy has shown that BiV can reduce activation delays between left ventricle (LV) and right ventricle (RV) (2). This imaging technique has the ability to quantify separately interventricular and intraventricular dyssynchrony, but the characteristics and the compared prognostic values of interventricular and left and right intraventricular dyssynchrony have not been previously fully described.
One hundred and three patients with idiopathic dilated cardiomyopathy (IDC) according to the definitions of the World Health Organization (3)were consecutively and continuously studied between January 1992 and June 2001. Diagnosis was established by normal coronary angiography in all patients, echocardiography, and equilibrium radionuclide-gated blood pool angiography (ERNA). Patients with chronic atrial arrhythmias, sinus node dysfunction, atrioventricular block, or with a permanent pacemaker were excluded. The patients were treated during follow-up with long-term angiotensin-converting enzyme inhibitors (86%), diuretics (70%), digoxin (53%), beta-blockers since 1998 (40%), and/or amiodarone (26%) if necessary. Two patients received a BiV pacemaker during the follow-up and were censored at the time of implantation.
A control group included 20 healthy subjects age 42 ± 14 years with no cardiovascular symptoms, a normal general examination, normal electrocardiogram (ECG), normal echocardiography, and coronary angiography in subjects age >50 years.
Electrocardiography and investigations
All the investigations were performed once at baseline. The QRS duration was measured on surface 12-lead ECGs from the first deflection of the QRS complex to the terminal isoelectric component of the complex. Left bundle branch block (LBBB) was present in 26 (25%) patients, left anterior hemiblock in 23 (22%), incomplete LBBB in 8 (8%, right bundle branch block (RBBB) in 2 (2%), and mean QRS duration was 113 ± 32 ms. The other investigations that were performed to allow a valuable prognostic study with multivariate analysis included a hemodynamic study and an evaluation of the risk of ventricular arrhythmia: two-dimensional and M-mode echocardiography was performed in accordance with the American Society of Echocardiography recommendations. Left ventricular cavity dimensions at end-systole and end-diastole were measured, and fractional shortening index was calculated according to the standard formula. Measurement of peak exercise oxygen consumption was measured in 74 patients who could adequately perform exercise testing. Left heart catheterization by the Judkins technique was performed in all patients with normal coronary angiography, and right heart catheterization was performed with measurements of cardiac index and right atrial, right ventricular, pulmonary artery, and pulmonary capillary wedge pressures. Twenty-four-hour ambulatory ECG recording was performed in all patients with grading of ventricular arrhythmias. The evaluation of heart rate variability in time domain and a recording of signal-averaged ECG were performed with methods described in previous studies from our group (4,5).
Red blood cells were labeled in vivo with 740 to 925 MBq of technetium-99m. The ECG was monitored continuously to ensure R-wave gating of the QRS complex. Elimination of ventricular premature beats was obtained with a window threshold of 20% around the mean RR interval during acquisition of projections. Extrasystolic and postextrasystolic cycles were excluded. Multigated equilibrium blood pool scintigrams were acquired at rest until the number of counts was 600,000 in the “best-septal” left anterior oblique projection to provide optimal RV and LV blood pool discrimination. The projection was gated with the ECG to get 16 frames spanning the cardiac cycle.
Scintigrams were acquired for each patient in sinus rhythm. The RV and LV regions of interest were acquired at end-diastole and end-systole for the respective ventricle. Regions of interest were drawn automatically by the computer with adjustments of border definition performed by an observer blinded to the state of conduction. After correction for background counts, LV and RV ejection fractions (EFs) were computed using the formula:where EDCis end-diastolic counts and ESCis end-systolic counts.
Nuclear phase imaging
Phase images were generated from the scintigraphic data using a commercially available computer program. The identical scintigraphic data used to generate RV and LV EFs were digitally processed to display the “phase” for each pixel overlaying the equilibrium blood pool and gated to the ECG R-wave. The phase program assigns a phase angle to each pixel of the phase image, derived from the first Fourier harmonic of time. The phase angle corresponds to the relative sequence and pattern of ventricular contraction during the cardiac cycle. Color-encoded phase images with corresponding histograms were generated for each patient. Scintigrams were intensity-coded for amplitude, the other parameter of Fourier first harmonic study. Phase images were generated for cardiac regions using a continuous color scale, corresponding to phase angles from 0° to 360°. Mean phase angles were computed for RV and LV blood pools as the arithmetic mean phase angle for all pixels in the ventricular region of interest. Mode was the angle with the highest value on the histogram of phases. Interventricular dyssynchrony was evaluated with the difference between LV and RV mean phase angles (RV-LV delay) and also with its absolute value (interventricular delay) considering that some patients were found to have a negative RV-LV delay. Intraventricular contractile synchrony in each ventricle was measured as the standard deviation (SD) of the mean phase angle for the RV and LV blood pools. Results were expressed in angle (°) and time in milliseconds (mean of the cardiac cycle duration during acquisition · angle/360).
All values are given as mean ± SD. Comparisons between groups were made using the Student ttest. Linear regression analysis was carried out to correlate quantitative variables. Survival curves were estimated by the Kaplan-Meier method, and curves were compared using the log-rank test. Major cardiac events included cardiac death and worsening of HF, leading to heart transplantation. The effects of all the parameters were studied with univariate and multivariate regression analysis (proportional hazards model). A forward stepwise model with a p value for entry of 0.05 was used. Statview 5.0 (Abacus Concepts, Berkeley, California) was used for statistical analysis.
In the control group, mean phase of LV was 345 ± 42 ms, SD of mean phase in LV was 22 ± 12 ms, mean phase of RV was 336 ± 33 ms, and SD of mean phase in RV was 30 ± 10 ms. The RV-LV delay was 7 ± 13 ms and interventricular delay was 15 ± 13 ms.
The general characteristics of the patients are summarized in Table 1, and the results of the phase analysis of ERNA in degrees (°) and in milliseconds are shown in Table 2. Compared to all the other subjects, patients with LBBB had an increased RV-LV delay (51 ± 66 ms vs. 16 ± 35 ms, p = 0.001) and an increased left intraventricular dyssynchrony (SD of LV mean phase 77 ± 41 ms vs. 50 ± 22 ms, p < 0.0001). According to QRS duration < or > 120 ms (whether the patient had LBBB or RBBB), we found a clear increase in SD of LV in mean phase, of interventricular delay and at a less degree of SD of RV mean phase in patients with QRS duration >120 ms (Table 2). As the results of phase analysis of ERNA expressed in degrees did not provide more pertinent information than results in milliseconds, particularly at the time of the analysis of survival, we only display in the results the statistical analysis about phase analysis expressed in milliseconds to avoid redundancy, although the complete analysis has been performed for both units.
Correlation of phase analysis of ERNA with the status of patients with IDC
The QRS duration was mainly related to SD of LV mean phase and to interventricular delay and to a lesser degree to SD of RV mean phase and to RV-LV delay (Table 3). Although they are not displayed in the tables, these correlations were similar when considering patients with QRS duration < or > 120 ms. Correlation coefficients between hemodynamic status and ventricular dyssynchrony expressed in milliseconds are shown in Table 4. Degradation in hemodynamic status (i.e., decrease in LVEF, cardiac index, peak oxygen uptake, increase in LV diameters) was associated with an increase in left and right intraventricular dyssynchrony. Parameters concerning LV (diameters, EF) were best related to left intraventricular dyssynchrony (Fig. 1). Cardiac index inversely correlated with both left and right intraventricular dyssynchrony. Peak oxygen uptake inversely correlated with right intraventricular dyssynchrony. An increase in interventricular dyssynchrony was not associated or was only poorly related (for interventricular delay) with a degradation of the functional or hemodynamic status, with a lower New York Heart Association (NYHA) functional class and a higher peak oxygen uptake when RV-LV delay increased.
Follow-up and prognostic value of phase analysis of ERNA in IDC
With a follow-up of 27 ± 23 months, 18 patients had a major cardiac event (7 cardiac deaths; 11 worsening of HF leading to heart transplantation). The SDs of the LV and RV mean phase and QRS duration were univariate predictors of cardiac event (all p < 0.0001), but RV-LV delay ⇓and interventricular delay were not (Table 5) (Figs. 2 and 3). ⇓⇓Among 13 demographic, hemodynamic, and arrhythmic markers that were univariate predictors of cardiac event, the only independent predictors (proportional hazards model) of a major cardiac event were an increased SD of LV mean phase (p = 0.0004), and an increased pulmonary capillary wedge pressure (p = 0.009) (Table 5).
We found in this study that intraventricular dyssynchrony evaluated with phase analysis of ERNA was an independent predictor of cardiac event in IDC and that the prognosis of patients with IDC was related to left intraventricular rather than to interventricular dyssynchrony.
An interesting finding is that the QRS duration seemed to reflect similarly interventricular and left intraventricular dyssynchrony as coefficients of correlation with QRS duration were 0.56 and 0.57 for interventricular delay and SD of LV mean phase, respectively. Moreover, the location of the dyssynchrony seemed similar whether patients had a marked increase in QRS duration or not. Concerning interventricular dyssynchrony, the interventricular delay (whether LV or RV was delayed compared to the other ventricle) was more related to QRS duration or to the degree of LV dilation than the delay of LV compared to RV, a value that was often positive but that could be negative in some cases. The positive correlation between peak VO2and RV-LV delay is surprising and may only have been due to hazard in view of the number of comparisons performed.
Dilated cardiomyopathy is associated with structural abnormalities of ventricular myocardium with both ventricular activation and mechanical contraction (6). The proposed mechanisms for global cardiac abnormalities in isolated LBBB are altered ventricular activation, delayed LV contraction, RV-LV dyssynchrony, decreased LV diastolic time, or abnormal septal motion, with decreased LV EF (7). The prognostic value of ventricular dyssynchrony using QRS duration measured on surface ECG has been known for several years and has been found in several studies (8,9). Improvements in hemodynamic status have recently been demonstrated during atrial synchronized BiV (1). At this time, selection of patients with HF candidate to BiV is usually done with measurement of QRS duration using surface ECG, and good responders to BiV may experience a significant reduction in QRS duration during BiV compared to bad responders (10). Although the mechanisms of the deleterious effect of a long QRS duration may be multiple, the fact that the strategy of resynchronization therapy began with a biventricular stimulation is a reflection that the first idea was to decrease the interventricular delay. Some preliminary findings suggest that BiV or LV pacing may also induce LV intraventricular resynchronization (11). That may also explain the favorable results that are very often found with LV pacing alone (12,13)or that are nearly similar when comparing LV pacing to BiV (14–17)in HF patients in acute studies or with a mid-term follow-up. However, the fact that RV dyssynchrony correlated with QRS width, some hemodynamic parameters, and even with cardiovascular events implies that at least some patients have a significant RV dyssynchrony, even if it is not at the same degree as LV dyssynchrony. Therefore, compared to BiV, using LV pacing alone may increase RV asynchrony, which raises the question whether this may adversely affect prognosis.
Some methods recently proposed to evaluate more precisely the myocardial dyssynchrony have used cardiac ultrasound (18); Doppler interventricular delay calculated as the difference between aortic and pulmonary pre-ejection delay (19); three-dimensional echocardiography (20); color kinesis (21); or tissue Doppler imaging (22), which was perhaps the most promising method. However, although these methods provide some very accurate information about the location and the degree of dyssynchrony in ventricles and about the effects of resynchronization therapy with BiV, none of them at this time have really evaluated the prognostic value of the parameters that have been proposed.
We consider that our results may provide some interesting data for all the techniques that could be proposed to evaluate ventricular dyssynchrony in dilated cardiomyopathy. It seems more pertinent to find reliable and noninvasive parameters that could evaluate intraventricular dyssynchrony (mainly in LV but also in RV) rather than parameters reflecting interventricular dyssynchrony, because we found that interventricular delay was not a prognostic factor in our series of patients. Moreover, no clear relationships were found between the degradation of the functional and hemodynamic status and the increase in the interventricular dyssynchrony. Although highly speculative, one could suggest that the aim of the resynchronization therapy could rather be to obtain a decrease in parameters of intraventricular dyssynchrony if one hopes to improve clinical status and affect prognosis positively. As QRS duration is a reflection of both inter- and intraventricular dyssynchrony, it is possible that QRS duration may also decrease with an intraventricular resynchronization.
The studies about phase analysis of ERNA to evaluate dyssynchrony in HF focused on resynchronization therapy with BiV or LV pacing and concerned short series of patients with observation of the beneficial effects of the therapy. Toussaint et al. (23)used phase analysis of ERNA to quantify short- and long-term impact of BiV on cardiac synchronization in 21 patients with drug refractory dilated cardiomyopathy and marked electrical dyssynchrony (QRS 180 ± 15 ms) with measurements at baseline, day 8 after BiV pacemaker implant, and after a follow-up of 12 months with BiV. Biventricular pacing reversed LV apex-to-base activation time, and increased LV and RV EF. The study by Kerwin et al. (2)showed that during BiV, their patients’ interventricular contractile synchrony improved overall. The degree of interventricular dyssynchrony present at baseline correlated with the magnitude of improvement in synchrony during BiV. The LV EF increased in all 13 patients in the Kerwin et al. (2)study during BiV and correlated significantly with improvement in RV/LV synchrony during BiV. Moreover, Le Rest and colleagues (24)investigated 17 patients with either ischemic or primary dilated cardiomyopathy in NYHA functional class III or IV who received an LV pacemaker. They performed three equilibrium-gated blood pool studies in each patient, one before pacing and two after pacer implantation (one with pacing on, and one after turning off the pacemaker). Phase analysis demonstrated a significant decrease of the interventricular phase shift (delta[pi]) with LV pacing. Clinical improvement was observed in patients with an initial positive delta(pi) that decreased with pacing and/or an initial LV phase SD >50° that decreased with pacing. This latter result is another indicator that the beneficial impact of LV pacing is not limited to interventricular delay. The intraventricular dyssynchrony may also be improved (or may at least be modified) with LV pacing. As for the echocardiographic studies previously mentioned, an evaluation of the prognosis of the dyssynchrony and of its location was not an objective of any of these studies and was not performed.
We found that the results of phase analyses expressed in milliseconds were more pertinent than those expressed in degrees, a result that might be surprising. The results in milliseconds have the advantage that they take into account the duration of the cardiac cycle. Considering that tachycardia (or decreased duration of RR cycle) is a recognized risk factor of events in these patients, one might have thought that the expression of dyssynchrony in milliseconds was a reflection of both dyssynchrony and tachycardia. However, tachycardia should have decreased the quantification of dyssynchrony expressed in ms compared to degrees and it should be noted that the bad prognosis was related to increased measurements in ms of dyssynchrony. Thus, the results expressed in milliseconds may be more accurate to evaluate the real consequences of each dyssynchrony (intra LV, intra-RV, or interventricular on the cardiac filling and cardiac outpout.
Intraventricular dyssynchrony evaluated with phase analysis of ERNA had more important correlation with severity of the hemodynamic status than interventricular dyssynchrony in IDC. The left intraventricular dyssynchrony was an independent predictor of cardiac event in this study. The prognosis of patients with IDC seemed related to intraventricular rather than to interventricular dyssynchrony. This may have some implications for resynchronization therapy in these patients.
- biventricular pacing
- ejection fraction
- equilibrium radionuclide-gated blood pool angiography
- heart failure
- idiopathic dilated cardiomyopathy
- left bundle branch block
- left ventricle
- New York Heart Association
- right bundle branch block
- right ventricle
- standard deviation
- Received March 27, 2002.
- Revision received July 17, 2002.
- Accepted August 1, 2002.
- American College of Cardiology Foundation
- Kerwin W.F.,
- Botvinick E.H.,
- O’Connell J.W.,
- et al.
- Richardson P.,
- McKenna W.,
- Bristow M.,
- et al.
- Fauchier L.,
- Babuty D.,
- Cosnay P.,
- Autret M.L.,
- Fauchier J.P.
- Dec G.W.,
- Fuster V.
- Aaronson K.D.,
- Schwartz J.S.,
- Chen T.M.,
- Wong K.L.,
- Goin J.E.,
- Mancini D.M.
- Yu C.M.,
- Lau C.P.,
- Tse H.F.,
- et al.
- Touiza A.,
- Etienne Y.,
- Gilard M.,
- Fatemi M.,
- Mansourati J.,
- Blanc J.J.
- Blanc J.J.,
- Etienne Y.,
- Gilard M.,
- et al.
- Auricchio A.,
- Stellbrink C.,
- Block M.,
- et al.
- Kass D.,
- Chen-Huan C.,
- Curry C.,
- et al.
- Butter C.,
- Auricchio A.,
- Stellbrink C.,
- et al.
- Stellbrink C.,
- Breithardt O.,
- Franke A.,
- et al.
- Kim W.Y.,
- Sogaard P.,
- Mortensen P.T.,
- et al.
- Godoy I.E.,
- Mor-Avi V.,
- Weinert L.,
- et al.