|Home | About | Journals | Submit | Contact Us | Français|
Prolonged inter-atrial conduction time (IACT) can be associated with abnormal left atrial (LA) function but has not been characterized in patients with left ventricular systolic dysfunction (LVSD) and a QRS interval >130 ms.
2D-Doppler echocardiography and tissue Doppler imaging (TDI) was performed in 41 patients with LVSD (LV ejection fraction 26±5%) and 41 similarly-age normal controls (NC). 2D measurements included LV volumes, ejection fraction and LA volumes for determination of LA emptying fraction and LA ejection fraction. IACT was defined as onset of P-wave to onset of the TDI-derived late diastolic (A') velocity at the lateral mitral annulus. 2D-Doppler measurements were re-assessed in LVSD patients 4 ± 2 months after cardiac resynchronization therapy (CRT).
IACT was longer in LVSD patients compared to NC (105±25 vs. 74±12 ms, p <.001);none of the NC had IACT> 100ms.In LVSD patients, IACT correlated modestly with measurements of LA volumes (r =.41-.48, all p <.009) but not with measurements of LA function. LVSD patients with IACT > 100 ms (n=20) prior to CRT had larger LA volumes and lower indices of LA function after CRT compared to the ≤ 100ms group. Significant reductions in LV end-systolic volume and increases in LV ejection fraction occurred in both groups after CRT.
TDI-derived IACT can be prolonged in patients with severe LVSD and a wide QRS interval. An IACT > 100 ms. can affect LA remodeling and function at early follow-up after CRT but does not influence the response in LV end-systolic volume or ejection fraction.
Prolonged interatrial conduction time (IACT) is associated with the development of atrial fibrillation and abnormal left atrial (LA) function. (1-4) IACT can be measured by placement of an invasive catheter in the coronary sinus to record an intracardiac electrocardiogram, 12-lead or signal-averaged electrocardiographic (ECG) P-wave duration, or by 2-dimensional-Doppler echocardiography including tissue Doppler imaging. (4-9) IACT can be also prolonged in patients with atrioventricular conduction delay (i.e., PR interval > 200 ms). (4, 10)
IACT measured by 2D-Doppler echocardiography and its association with indices of LA function has been reported in a few studies of patients with left ventricular systolic dysfunction (LVSD) (2-8) but has not been investigated in patients with prolonged QRS interval (i.e., > 130 ms). In LVSD patients that receive cardiac resynchronization therapy (CRT). IACT has been reported to be an important consideration when programming the optimal atrioventricular delay (AVD) in patients that require atrial pacing. (5) The importance of IACT has not been studied prior to, or after CRT, during atrial-sensed biventricular pacing.
The objectives of this study were: 1) to evaluate the relation between IACT and LA function in patients with LVSD and prolonged QRS interval compared to normal controls; and 2) to determine the effects of CRT, at an optimal programmed AVD, and the relationship of IACT to LA and LV function in patients that did not require atrial pacing.
This was a retrospective, single-center study of patients with LVSD (i.e., LV ejection fraction < 35%), medically refractory NYHA Class III or IV heart failure symptoms, and a QRS interval > 130 ms. None of the patients required atrio-ventricular sequential pacing prior to or following CRT. A group of age-matched Normal Controls, previously evaluated for a separate research protocol, with no symptoms or cardiovascular risk factors, normal blood chemistries, normal blood pressure (determined in both arms based on the average of 3 recordings), a normal 12-lead ECG, and a negative symptom-limited exercise stress echocardiogram, were included for comparison. The study protocol was approved by the Human Research Protection Office at Washington University School of Medicine, St. Louis, Missouri.
A standard 12-lead ECG and a 2-dimensional (2D) pulsed-wave Doppler (PWD) echocardiographic examination with tissue Doppler imaging were performed in all participants. ECG measurements included the maximal in any lead of the P-wave duration, PR interval, and QRS duration. 2D measurements included LV volumes and calculation of ejection fraction, LA diameter in the parasternal long-axis view, and LA volumes in the apical 4-chamber view by the method of discs. LA emptying and systolic ejection fraction was calculated as in prior studies from this laboratory. (11) The LA emptying fraction was determined as (LA maximal volume at end-systole - LA minimal volume at end-diastole) / LA maximal volume at end systole. LA systolic ejection fraction was determined as (LA volume at P-wave onset - LA minimal volume at enddiastole) / LA volume at P wave onset (Figure 1). Color flow imaging in the parasternal long-axis and the apical 4-chamber view was performed for measurements of jet width at the vena contracta in patients with greater than mild mitral regurgitation (i.e., ≥ 4.0 mm). (12)
PWD-derived transmitral inflow velocities were obtained at the mitral leaflet tips in the apical 4-chamber view for measurements of peak early (E-wave) filling velocity, late diastolic velocity (A-wave), and A-wave duration. Tissue Doppler imaging was performed in the apical 4-chamber view with a 2.5 mm sample volume placed in the lateral mitral annulus. Measurements included the peak early-diastolic (E') and late-diastolic mitral annular velocity (A'). The ratio of mitral E-wave to E' velocity was derived as an estimation of LV filling pressure. IACT was determined as the time from onset of P-wave to the onset of the tissue Doppler-derived A-wave (P-A') velocity (Figure 2). All Doppler spectral velocities were recorded at 100 mm/s sweep speed. 2D-Doppler echocardiographic measurements were acquired in 3-5 cardiac cycles and averaged.
The interobserver and intraobserver variability for measurements of the P-A' interval was determined in 15 normal controls and 15 LVSD patients. Correlation coefficients (Pearson's r value) for interobserver variability were 0.82 (difference: 4±4 ms) and 0.80 (difference: 9±9 ms) for normal controls and LVSD patients, respectively; intraobserver variability was 0.87 (difference: 4±3 ms) and 0.89 (difference: 8±6 ms) for normal controls and LVSD patients, respectively.
CRT device implantation in the LVSD patients was performed by the transvenous approach. The right atrial lead was positioned in the appendage and the right ventricular lead was positioned in the apex or ventricular septum. The LV lead was placed in the mid-lateral coronary vein when possible; anterolateral or posterolateral coronary vein LV lead placement were used if the lateral branch was either absent or unacceptable because of phrenic nerve stimulation or high stimulation threshold. After implantation, the CRT device was programmed to ventricular demand pacing (VVI) at a rate of 40 beats per minute (i.e., no CRT) and the baseline echocardiographic study was performed within 24 hours. The optimal AVD was determined by the Doppler-derived aortic velocity time integral. (13, 14) All CRT devices were programmed to simultaneous biventricular pacing and the atrial-paced offset was set at 40ms. 2D-Doppler echocardiography and TDI was repeated at 4 ± 2 months after CRT; the programmed AVD was not changed during follow-up.
Continuous variables are expressed as the mean ± 1 standard deviation. Comparisons of echocardiographic variables between groups (i.e., Normal control vs. LVSD) and within LVSD groups (i.e., Pre-CRT vs. Post-CRT) were performed using paired Student's t-test. Pearson correlation coefficients were determined between continuous variables and Fisher's exact test for comparison of categorical variables when appropriate. Statistical analyses were performed using SAS (version 9.1, SAS Institute, Cary, North Carolina, USA); significance was defined as a p < .05.
The characteristics of the study population are shown in Table 1. Patients with LVSD had higher heart rates and longer PR intervals compared to Normal controls; 19 (46%) had a PR interval > 200 ms. P-wave duration did not significantly differ between groups; only 11 (27%) patients with LVSD had a P-wave duration ≥ 120 ms.
2D-derived LV and LA volumes were larger and indices of LA function (i.e., LA emptying and ejection fraction) were significantly lower in the LVSD group compared to Normal controls. The PWD-derived transmitral A-wave velocity was higher and A-wave duration was significantly longer in the LVSD group. The TDI-derived A' velocity was lower and the mitral E/E' was higher in the LVSD group. IACT was significantly longer in the LVSD group. The median P-A' interval was 100 ms. None of the Normal controls had a P-A' > 100 ms.
In the LVSD group, IACT correlated with P-wave duration (r = .46, p = .002), PR interval (r = .73, p < .0001), and QRS duration (r = .44, p = .004). The IACT correlated with LA end-systolic volume (r = .43, p =.006), LA volume prior to atrial systole (r = .41, p =.008), and LA end-diastolic volume (r = .48, p =.002), but not with measurements of LA function or LV ejection fraction.
In the Normal control group, IACT correlated with the PR interval (r = .43, p < .01), but not with P-wave (r = .25) or QRS (r = -.11) duration. IACT correlated weakly with the LA volume prior to atrial systole (r = .31, p < .05), and with LA end-diastolic volume (r = .32, p < .05), but not with measurements of LA function or LV ejection fraction.
Patients with LVSD were grouped according to the median P-A' interval of 100 ms to determine whether IACT influenced the response in LV and/or LA function after CRT (Table 2). The IACT > 100 ms group (n = 20) were older and had significantly longer P-wave duration, PR and QRS intervals.
LV volumes were smaller in the IACT > 100 ms group prior to CRT. However, LV endsystolic volume decreased significantly and LV ejection fraction increased significantly in both groups after CRT (Table 3). In the IACT > 100 ms group, the LA emptying fraction decreased and the mitral E/E' ratio increased after CRT. Conversely, in the IACT ≤ 100 ms group, there were no significant changes in IACT, LA function, or the mitral E/E' ratio after CRT. Mitral regurgitant jet width in patients with moderate to severe regurgitation (n=25) decreased after CRT in both groups but did not achieve statistical significance. After CRT, LA volumes remained larger, measurements of LA function were lower and the mitral E/E' ratio was higher in the IACT > 100 ms group compared to the IACT ≤ 100 ms group. Although IACT decreased in the > 100 ms group after CRT, it remained significantly longer compared to the ≤ 100 ms group (p = .03). The results suggest that a longer IACT prior to CRT may have influenced the effects of CRT on LA remodeling. However, when patients were dichotomized based on reduction in LV endsystolic volume > 15% after CRT (n=25, 61%) there were no differences in the results with regards to the effect on IACT, LA volumes or function.
The programmed AVD for CRT was shorter than the IACT in 14/41 patients (34%). This was noted more often in the IACT > 100 ms group compared to the IACT ≤ 100 ms group (10/20 [50%] vs. 4/21 [19%], respectively, p = .04). However, there were no significant differences in the response in LV end-systolic volume, LV ejection fraction, LA volume, LA function, or LV filling pressures after CRT between the two groups (data not shown).
The results of this study reveal several findings regarding IACT determined by tissue Doppler imaging in patients with LVSD and prolonged QRS interval. First, although IACT correlated with measurements of 12-lead ECG determined P-wave duration, the correlation was stronger with the PR interval (i.e. AV conduction). Second, IACT correlated modestly with LA volumes but not with measurements of LA function. Third, a longer IACT (i.e., > 100 ms) prior to CRT was associated with larger LA volumes, worse LA function, and higher LV filling pressures after CRT, but did not affect the response in LV end-systolic volume or ejection fraction. Fourth, although the optimal programmed AVD for CRT was shorter than the IACT in 1/3 of patients, the response in LV ejection fraction was similar at early follow-up after CRT. Thus, a prolonged IACT may affect LA remodeling and function at follow-up after CRT but does not influence the response in LV end-systolic volume or LV ejection fraction.
Prolonged interatrial conduction is often present in patients with an increased P-wave duration and/or a prolonged PR interval. (4, 10) The 12-lead ECG measurements of P-wave duration did not significantly differ between the LVSD group and Normal controls despite significant differences in the TDI-derived IACT and PR intervals. A recent study also reported that P-wave duration did not significantly differ in a group with LVSD compared to healthy controls, although the PR and QRS intervals were not reported. (8) In the present study, IACT correlated only modestly with P-wave duration, and to a greater extent with the PR interval in patients with LVSD. It is known that prolonged AV conduction can be present in patients with LVSD. and a wide QRS interval (17). Our results are consistent with this prior report and the PR interval was an important variable with regards to interatrial conduction time (i.e., P-wave duration) in patients with LVSD and a wide QRS interval. Furthermore, TDI-derived IACT correlated modestly with LA volumes but there were no significant correlations with 2D-Doppler measurements of LA function or LV ejection fraction.
In this study, the median IACT was 100 ms in patients with LVSD. None of the Normal controls had a P-A' > 100 ms. Although the P-A' interval shortened in the IACT > 100 ms after CRT, there were significant decreases in the LA emptying fraction and increases in LV filling pressures (i.e., mitral E/E' ratio). Prior to CRT, this group had a longer P-wave duration, PR interval, and QRS interval, suggesting greater abnormalities in interatrial, atrioventricular, and interventricular conduction. After CRT, LA volumes were larger and LA function remained significantly lower in the IACT > 100 ms group, compared to the IACT ≤ 100 ms group. The increase in LV filling pressures in the IACT > 100 ms group after CRT may have resulted from the larger LA volumes and lower indices of LA function. However, LV end-sytolic volumes decreased and LV ejection fraction increased significantly in both groups after CRT. This is consistent with a prior study that reported the IACT prior to CRT was not a factor in the response in LV end-systolic volumes after CRT. (9)
There are few reports regarding the effect of CRT on LA volume and function. In the VIGOR-CHF trial, LA end-systolic volume decreased after CRT. (18) However, the results of the present investigation and another study (19), did not demonstrate a decrease in LA end-systolic volume after CRT. It has been reported that LA emptying fraction increased after CRT, which would suggest an improvement in LA function. (19) However, the present investigation did not demonstrate any improvement in LA function after CRT. This may be related to a lack of change in severity of mitral regurgitation (i.e., jet width), LV filling pressures (i.e., E/E') were only modestly elevated prior to CRT and remained unchanged after CRT, or to differences between patients in the present study compared to results of the prior investigation. The results of CRT on LA function also did not differ when patients were grouped based on reduction in LV end-systolic volume > 15% after CRT. These divergent findings suggest further studies are needed regarding the effect of CRT on LA function.
The programmed AVD for CRT, determined by the aortic VTI method, was shorter than the IACT in 34% of the patients. It has been reported that a programmed AVD shorter than the intrinsic atrial conduction time may result in a reduction in LV systolic performance. (5, 10) However, the response in LV ejection fraction, LA function, and LV filling pressures after CRT did not significantly differ in patients with a programmed AVD shorter than the IACT compared to patients with a programmed AVD longer than the IACT. These results suggest that during atrial-sensed biventricular pacing, the IACT, relative to the programmed AVD does not have a significant effect on the response in LV or LA function during CRT.
The number of patients with LVSD reported in this investigation was small and follow-up after CRT was confined to 3-6 months. This may have limited the power of the study to detect differences. No measurements of IACT were performed by intracardiac electrograms or by signal-averaged 12-lead ECG measurements of P-wave duration for comparison to the TDI method. However, it has been reported that TDI measurements of IACT correlate with signal-averaged P-wave duration. (7) By design, none of the patients in this study were atrial-paced prior to or following CRT. It is known that atrial pacing prolongs IACT and can affect the response in LV systolic performance during CRT. (5, 20) Thus, the results of this study may not be applicable to patients that require atrial pacing during CRT.
TDI-derived IACT can be prolonged in patients with severe LVSD and a wide QRS interval. Prior to CRT, a prolonged IACT > 100 ms. can affect LA remodeling and function at early follow-up after CRT but does not influence the response in LV end-systolic volume or ejection fraction.
Funding Supported in part by grants from the American Society of Echocardiography to A.D.W., National Institute of Health grants K12RR023249 and KL2RR024994 to L.d.l.F., and from the Barnes-Jewish Hospital Foundation to the Cardiovascular Imaging and Clinical Research Core Laboratory
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Disclosures ADW is a consultant for Boston Scientific Corporation and St Jude Medical Inc. VGD is a consultant for St. Jude Medical Inc. All other authors have no relationships to disclose.