Long-term right ventricular apical pacing (RVAP) can lead to adverse clinical outcomes. Although left ventricular (LV) dyssynchrony is the major causative factor, other potential mechanisms are not fully understood. We sought to clarify whether RVAP elicits apical wall motion abnormalities that contribute to LV contractile dysfunction.
We studied annual echocardiographic data over a 5-year period after pacemaker implantation (PMI) for 74 patients who underwent RVAP. The patients were divided into two groups according to the percentage of ventricular pacing: right ventricular (RV) pacing <50% and RV pacing ≥50%. We assessed LV ejection fraction, LV end-diastolic volume, and left atrial dimension. To assess regional wall motion abnormalities, the wall motion score index was calculated.
LV wall motion abnormality was observed in 64% of the subjects and was more pronounced in apical segments than in other segments. At 2 years after PMI, brain natriuretic peptide levels were significantly higher in the group with RV pacing ≥50% than in the group with RV pacing <50%. The subjects with RV pacing ≥50% had higher LV end-diastolic dimension and lower ejection fraction at 3 years after PMI.
Long-term RVAP elicits apical wall motion abnormalities that could in part contribute to LV contractile dysfunction.
Cardiac pacing is a safe, effective therapy for patients with bradyarrhythmia. Right ventricular apical pacing (RVAP) is a standard, widespread procedure owing to the accessibility of the right ventricular (RV) apical site, which provides stability for the pacing device. Recent clinical studies have reported that RVAP elicits abnormal left ventricular (LV) contractions and reduced pump function. Pacing from the RV apex can alter the electrical LV activation sequence and lead to dyssynchrony in LV contraction, resulting in impaired hemodynamic function. Many studies have examined the detrimental effects of RVAP, and a number of investigators have focused on the relationship between the pacing site and intraventricular dyssynchrony. LV dyssynchrony is a critical factor in LV dysfunction; however, other potential mechanisms are not fully understood.
So far, little attention has been given to regional myocardial damage at the pacing site. Previously, Tl-201 scintigraphic studies in humans showed that long-term RVAP resulted in a high incidence of myocardial perfusion defects in the LV apical region. In an experimental investigation using dogs, interstitial fibrosis and extracellular matrix remodeling, which are the cardinal features of tissue remodeling in heart failure, were obvious in apical pacing models. These results led to the hypothesis that electrical stimulation delivered at the right apical site damages the regional LV apical myocytes and produces intraventricular dyssynchrony, resulting in wall motion abnormalities around the pacing site and global LV contractile dysfunction. However, it is difficult to conduct experimental animal studies involving long-term observations of the surrounding myocardium. The present investigation sought to clarify whether RVAP elicits abnormal LV apical wall motion leading to LV contractile dysfunction.
Materials and Methods
This study retrospectively reviewed 86 consecutive patients who underwent pacing treatment between April 2000 and March 2002 at Akita University Hospital. The patients had sick sinus syndrome (SSS), complete atrioventricular block (AVB), or atrial fibrillation with bradycardia. All patients agreed to pacemaker implantation (PMI) and provided written informed consent to the study protocol, which was approved by the clinical research and ethics committee of the University of Akita.
The enrolled patients were experiencing symptoms related to bradyarrhythmia and had preserved cardiac function with an LV ejection fraction (LVEF) of ≥50%. The patients excluded from this study were those with 1) ischemic heart disease ( n = 7); 2) an LVEF <50%; 3) cardiomyopathies such as cardiac sarcoidosis ( n = 2), cardiac amyloidosis ( n = 1), hypertrophic cardiomyopathy, or dilated cardiomyopathy; 4) implantation of a resynchronization pacemaker or cardiac defibrillator ( n = 2); 5) implantation of a non-apical pacemaker; or 6) a changing atrioventricular delay and pacing stimulation threshold. Seventy-seven of the 86 patients (89.5%) had confirmed coronary angiography to exclude ischemic heart disease. Because of chronic kidney disease, 9 patients could not undergo coronary angiography, but instead had thallium-201 exercise myocardial scintigraphy. Coronary artery disease was defined as ≥75% stenosis in at least one of the major coronary arteries. A total of 12 patients were excluded. Consequently, a total of 74 patients were included in the final analysis.
After PMI, all patients were examined at Akita University Hospital during outpatient visits and observed every 2 months for 5 years. Adverse events were defined as death from any cause and a first occurrence of hospitalization for congestive heart failure. The diagnosis of congestive heart failure was validated using Framingham criteria.
PMI followed standard procedures using a transvenous approach via the left infraclavicular route. The RV leads were positioned at the RV apex, and the right atrial leads were placed at the right atrial appendage. The stimulation threshold in the RV leads was <1.5 V, with a pulse width of 0.4 ms; the intracardiac R wave was >5 mV. The parameters for right atrial lead positioning were threshold <1.5 V, pulse width 0.4 ms, and P wave >3 mV. VDD mode pacemakers were implanted in patients with AVB, DDDR mode pacemakers were implanted in patients experiencing SSS and AVB, and VVIR mode pacemakers were implanted in patients with long-lasting atrial fibrillation, as defined previously. All pacemakers were programmed at a lower rate of 60 bpm, an upper rate of 120 bpm, and a paced atrioventricular delay of 180 msec. The atrial and ventricular pacing stimulation voltage was three times the threshold stimulation voltage, and these settings were not changed during follow-up.
Two-dimensional echocardiography was performed using a SEQUOIA 512 (Siemens, Erlangen, Germany). Echocardiographic images were obtained in the standard parasternal long and short axes with apical four-, two-, and three-chamber views. For wall motion assessment, at least four cardiac cycles were acquired at the LV base, mid-papillary muscle level, and apex. The presence of an apical wall motion abnormality was recognized if it was observed in both the parasternal short-axis view and apical four-, two-, or three-chamber view. Regional wall motion was assessed according to the standard 16-segment model recommended by the American Society of Echocardiography. The wall motion score index (WMSI) was calculated for regional wall motion analysis. In this scoring system, a higher score indicates a more severe wall motion abnormality (1, normal; 2, hypokinesis; 3, akinesis; 4, dyskinesis; and 5, aneurismal). The WMSI was calculated by dividing the sum of the wall motion scores by the number of segments visualized:
WMSI = Sum of wall motion scores Number of visualized segments
The LVEF and LV end-diastolic volume (LVEDV) were measured using Simpson’s biplane method. All echocardiographic measurements were performed with full pacing of the ventricle. Echocardiographic analyses were performed before PMI, immediately after surgery (before discharge from the hospital), and at 1, 2, 3, and 5 years after PMI. Experienced sonographers analyzed the echocardiographic images, which were stored on Super-VHS. Two independent echocardiologists who were blind to the clinical data reviewed all images off-site. LV dyssynchrony was assessed at baseline and at 5 years after PMI. In addition to the paced QRS duration obtained from 12-lead electrocardiograms, the septal-to-posterior wall motion delay (SPWMD) was measured using M-mode recordings from the parasternal short-axis view. Specifically, the interval between the maximal posterior displacement of the septum and maximal displacement of the LV posterior wall was measured.
The interobserver variability for all measurements and regional wall motion scores was analyzed by finding the difference between two measurements made by two observers. The intraobserver variability was analyzed using two measurements by one observer. Inter- and intraobserver agreements for these measurements were evaluated using intraclass correlation coefficients. The following values for intraclass correlation coefficients (±95% CI) were obtained, suggesting good reproducibility of the echocardiographic parameters. Interobserver variability was as follows: LVEF 0.949 (0.878–0.979), LVEDV 0.906 (0.782–0.961), left atrial dimension 0.874 (0.714–0.948), and WMSI 0.973 (0.933–0.989). Intraobserver variability was as follows: LVEF 0.974 (0.936–0.989), LVEDV 0.933 (0.833–0.973), left atrial dimension 0.974 (0.955–0.986), and WMSI 0.987 (0.967–0.995).
Continuous variables are expressed as the mean ± SD. For continuous and normally distributed data, Student t test was used to compare groups. For non-normally distributed data, the Mann–Whitney U test was used. A multiple linear regression analysis was performed to estimate factors influencing the LVEF at 5 years after PMI. All parameters with a significance value less than .1 by univariate analysis were entered into the multivariate model. Cumulative event rates were calculated using the Kaplan–Meier method. Comparisons among groups were made using log-rank tests. A P value < .05 was considered statistically significant. All statistical analyses were performed with SPSS for Windows Ver. 13.0 (SPSS, Chicago, IL).
Baseline Patient Characteristics
A total of 74 patients were included in the final analysis. The baseline characteristics of the study population are summarized in Table 1 . The patients were divided into two groups according to percentage of ventricular pacing: RV pacing <50% ( n = 34) and RV pacing ≥50% ( n = 40). The average ventricular pacing for each patient was calculated from all stored data from pacemaker device interrogations over 5 years. At 5 years after PMI, the patients were divided into two groups according to the average of ventricular pacing. The distribution of all patients is given in Supplementary Figure 1 . The pacing and sensing rates were calculated automatically by the individual devices. There were no significant differences in mean age, sex, prevalence of underlying heart disease, drug use, or echocardiographic parameters between the groups ( Table 1 ). The prevalence of SSS was significantly higher in the group with RV pacing <50% (94.1%, n = 32; P < . 01), whereas the prevalence of AVB was significantly higher in the group with RV pacing ≥50% (60.0%, n = 24; P < . 01). There was a significant difference in the percentage of ventricular pacing between the group with RV pacing <50% and the group with RV pacing ≥50% (12.0%± 11.5% vs. 92.0% ± 10.4%, respectively; P < . 01). Almost all of the patients were programmed with the DDDR pacing mode ( n = 68, 91.9%). Only two patients (2.7%) were programmed with the VDD pacing mode, and four patients (5.4%) were programmed with the VVIR pacing mode.
|All patients |
( n = 74)
|RV pacing <50% |
( n = 34)
|RV pacing ≥50% |
( n = 40)
|Age, y||69.3 ± 13.3||68.7 ± 12.0||69.9 ± 14.3||.69|
|Male sex, n (%)||37 (50.0)||16 (47.0)||21 (52.5)||.65|
|Sick sinus syndrome, n (%)||48 (64.9)||32 (94.1)||16 (40.0)||<.01|
|AVB, n (%)||26 (35.1)||2 (6.9)||24 (60.0)||<.01|
|Acute heart failure||41 (55.4)||18 (53.0)||23 (58.0)||.71|
|DDDR||68 (91.9)||33 (97.1)||35 (87.5)||.14|
|VDD||2 (2.7)||0 (0)||2 (5)||.16|
|VVIR||4 (5.4)||1 (2.9)||3 (7.5)||.38|
|Percentage of atrial pacing, %||63.0 ± 34.0||70.0 ± 40.4||57.1 ± 26.6||.12|
|Percentage of ventricular pacing, %||55.3 ± 41.6||12.0 ± 11.5||92.0 ± 10.4||<.01|
|Underlying heart disease, n (%)|
|Hypertension||42 (73.7)||19 (55.9)||23 (58.0)||.89|
|Valvular heart disease||55 (74.3)||25 (74.0)||30 (75.0)||.89|
|Medication, n (%)|
|ACEIs/ARBs||41 (55.4)||26 (76.5)||15 (37.5)||.28|
|Statins||15 (20.2)||7 (20.5)||8 (20.0)||.95|
|Left atrial diameter (mm)||41.4 ± 8.6||41.0 ± 7.5||41.8 ± 9.4||.69|
|LVEDV (mL)||107.4 ± 3.7||106.4 ± 6.6||108.1 ± 3.9||.82|
|Ejection fraction (%)||66 ± 8||65 ± 8||66 ± 8||.53|
Although all patients had no wall motion abnormalities before PMI, LV contractile dysfunction was seen in 64% of the patients at 5 years after PMI. In Figure 1 , the changes in echocardiographic parameters are compared between the groups. No differences in LVEF were observed between the groups for up to 2 years after PMI; however, a difference was noticeable at 3 years (63% ± 9% vs. 59% ± 10%, RV pacing <50% vs. RV pacing ≥50%; P = .038). In the group with RV pacing ≥50%, the LVEF decreased significantly, from 66% ± 8% at baseline to 53% ± 7% at 5 years ( P < . 001), but not in the group with RV pacing <50% (65% ± 8% vs. 63% ± 7%, baseline vs. 5 years; P = .306). At 5 years, there was a significant difference in LVEF between the groups ( P < . 001). The LVEDV was significantly increased at 3 years (143.8 ± 4.2 mL) compared with baseline (108.1 ± 3.9 mL; P < . 001) in the group with RV pacing ≥50%, but not in the group with RV pacing <50% (106.4 ± 6.6 vs. 109.9 ± 5.8 mL, baseline vs. 3 years; P = .128). The LVEDV differed significantly between the two groups from 3 years until the end of the follow-up period. No obvious differences in left atrial size were found between the groups. At 5 years after PMI, we did not find obvious differences in SPWMD between the groups (93.0 ± 8.9 vs. 100.2 ± 18.3 msec, RV pacing <50% vs. RV pacing ≥50%; P = .73). No differences in paced QRS duration were observed between the groups (154.6 ± 7.80 vs. 170.1 ± 4.43 msec, RV pacing <50% vs. RV pacing ≥50%; P = .069).