Background
Endovascular stenting has emerged as a treatment option for children with coarctation of the aortic (CoA), but the impact on left ventricular (LV) function has been poorly documented. The aim of this study was to characterize the LV myocardial and hemodynamic response to exercise in young patients who underwent endovascular stenting for CoA during childhood using semisupine bicycle exercise stress echocardiography.
Methods
This was a single-center prospective cross-sectional study including 30 patients with CoA and 30 age- and gender-matched control subjects who underwent semisupine bicycle exercise stress echocardiography. Color Doppler tissue imaging peak systolic (s′) and early diastolic (e′) velocities in the LV lateral wall and basal septum, LV myocardial acceleration during isovolumic acceleration were measured at rest and with incremental heart rate (HR). The relationship with increasing HR was evaluated for each parameter by plotting the values at each stage of exercise versus HR.
Results
At rest, HR was similar between the two groups. LV ejection fraction and fractional shortening were within the normal range in the CoA group. LV lateral wall and basal septal s′ and e′ velocities did not differ between the two groups, but isovolumic acceleration values were significantly lower in the CoA group. At peak exercise, HR was similar between the groups, but all Doppler tissue imaging parameters were lower in patients than in control subjects. When assessing the increase of each parameter versus HR, the increase in slope was significantly lower in patients than in control subjects for LV lateral wall Doppler tissue imaging s′ and e′ velocities, and septal e′ velocity, but not for septal s′. The relationship of isovolumic acceleration with HR was significantly reduced in the CoA group.
Conclusion
The results of this study demonstrate reduced systolic and diastolic myocardial reserve in patients with CoA compared with control subjects. An abnormal myocardial contractile response to exercise was also found, as shown by an abnormal LV force-frequency relationship in patients with stented CoA. The prognostic clinical implications require further study.
Coarctation of the aorta (CoA) is a common lesion, accounting for 5% to 8% of all congenital heart defects. Symptomatic neonates are treated surgically, with good outcomes. For children diagnosed beyond the newborn period (>3 months), balloon angioplasty and, in older children, stent implantation have evolved as treatment modalities with favorable short- and intermediate-term outcomes. Despite successful initial surgical or percutaneous management, CoA is a chronic condition with long-term morbidities related to residual arch obstruction, recoarctation, arterial hypertension, early coronary artery disease, and heart failure. Endovascular stenting is used primarily to treat native or residual CoA in older children and adults. Because of the late diagnosis or residual or recurrent aortic narrowing, the left ventricle in this patient group has been exposed to a prolonged increased pressure loading, which increases the risk for developing left ventricular (LV) dysfunction. In an earlier study, we evaluated the effect of stent implantation on blood pressure (BP) and LV hypertrophy and function at rest, and we demonstrated preserved or increased LV ejection fraction (LVEF) at rest. In the present study, we wanted to explore the hemodynamic and LV myocardial response to exercise in the same patient cohort. We hypothesized that the LV myocardial systolic and diastolic response would be reduced because of subclinical myocardial damage related to chronic pressure loading. Doppler tissue imaging (DTI) and strain imaging have been used to study myocardial response to exercise in patients with congenital heart disease. The aim of this study was to evaluate the LV myocardial response to exercise as measured by DTI in young patients after stent implantation for CoA using semisupine bicycle exercise stress echocardiography.
Methods
This was a single-center, prospective, cross-sectional study of patients who had undergone stent implantation for CoA during childhood. A search of the cardiac interventional database at the Hospital for Sick Children was performed to identify patients who had undergone stent implantation for CoA at <18 years of age, between September 1995 and November 2009. Indications for stent implantation were a cuff systolic arm-to-leg BP gradient > 20 mm Hg in addition to an angiographically confirmed lesion, either recurrent CoA (reCoA) or native CoA (naCoA), in the isthmus region of the aorta. Procedural success was defined using an arm-to-leg systolic BP (SBP) gradient ≤ 20 mm Hg, a normal abdominal aortic pulse Doppler flow profile, and a Doppler pressure gradient ≤ 20 mm Hg across the aortic isthmus. Once written consent was obtained, patients were recruited to participate in the study, which included clinical and echocardiographic assessment, 24-hour ambulatory BP monitoring, and semisupine bicycle exercise stress echocardiography.
The findings obtained in patients with CoA were compared with those from age- and gender-matched control subjects. Control subjects were selected from those with normal results on echocardiographic studies performed for the evaluation of a heart murmur, chest pain, palpitations, and/or syncope or from healthy volunteers. The institutional research ethics board approved the study.
Clinical Data
At the time of study, the ages, gender, heights, and weights of all participants were recorded. Resting heart rate (HR) and BP and resting arm-to-leg BP were measured. BP was measured with a GE Dinamap ProCare system (Critikon, Tampa, FL); resting BP was recorded as an average of two readings taken from the right arm while sitting, over a 5- to 10-min period before exercise echocardiography. Hypertension was defined as resting SBP or diastolic BP > 95th percentile for gender and height in pediatric patients (<18 years of age) and SBP ≥ 140 mm Hg or diastolic BP ≥ 90 mm Hg in adults (≥18 years of age). All participants had clinically palpable femoral pulses.
Semisupine Bicycle Ergometry
Semisupine bicycle exercise stress echocardiography was performed on a cycle ergometer (Lode BV, Groningen, The Netherlands) using a modified Bruce protocol. The ergometer resistance was increased by 20 W every 3 min in subjects ≤14 years of age and by 25 W in subjects >14 years of age. The test was interrupted either for symptoms (fatigue, chest pain, electrocardiographic changes) or when the target HR (80% of the maximal HR for age) was reached. During exercise, images were recorded at every 10 to 15 beat/min increase in HR while pedaling. At each stage, image loops of ≥10 beats were captured to ensure sufficient images for offline analysis. During exercise, peak gradients were also recorded through the stent using a continuous-wave Doppler technique.
Echocardiographic Acquisition and Analysis
Echocardiography was performed using a Vivid 7 echocardiographic system (GE Medical Systems, Milwaukee, WI). A full resting echocardiographic examination was performed before semisupine bicycle exercise stress echocardiography using a standard clinical protocol, in accordance with published guidelines. At rest and peak exercise, peak flow velocity across the aortic stent was measured from continuous-wave Doppler recorded from the suprasternal view. Pulsed-wave Doppler was used to evaluate flow velocity just proximal to the aortic stent. The corrected flow velocity across the aortic stent was obtained by subtracting proximal velocity from the peak flow velocity across the stent. At each stage of exercise, we obtained color Doppler images of the LV lateral wall and the interventricular septum from the apical four-chamber view. To optimize frame rates, the sector was narrowed for each wall. For all views, at least three complete cardiac cycles were recorded and stored in raw Digital Imaging and Communications in Medicine format for off-line analysis.
Resting and exercise echocardiographic parameters were analyzed offline using EchoPAC software (GE Medical Systems). All conventional systolic and diastolic parameters were measured according to published guidelines. Fractional shortening (FS) was measured from the parasternal short-axis view by M-mode imaging. LVEF was determined using the biplane Simpson method. Indexed LV mass (LVMi) was determined using the Devereux formula. Color DTI was performed using a 5-mm sample volume placed in the middle of the myocardium at the basal third of the LV free wall and basal septum. During analysis, manual tissue tracking was used to ensure that the sample volume remained within the myocardium throughout the cardiac cycle. The resulting Doppler spectral trace was displayed, and data points were smoothed with a three-sample average. Peak systolic and early diastolic DTI velocities were measured at each stage of exercise. Fusion of the e′ and a′ waves was frequently seen during exercise, and this fused wave was measured as e′. Velocity measurements were recorded as the average value from three consecutive cardiac cycles. Isovolumic acceleration (IVA) was obtained as described previously ( Figure 1 ). At each HR, three measurements of IVA were made, and an average was recorded.
At rest, LV longitudinal strain measurements were performed using 2D speckle-tracking echocardiography. The result was presented as a global value of LV longitudinal strain, defined as the arithmetic mean value of segmental longitudinal strain indices obtained in the basal, mid, and apical segments of septal and lateral walls of the left ventricle.
Reproducibility
To assess interobserver variability, two observers independently performed offline analysis of exercise echocardiographic parameters in 15 randomly selected participants. To assess intraobserver variability, the primary observer performed repeated measures of echocardiographic parameters during exercise at each HR in these 15 studies 2 weeks after the first analysis.
Statistical Analysis
Descriptive statistics of continuous data are presented as mean ± SD, median (interquartile range [IQR]), or minimum and maximum values, as appropriate. Comparisons between control subjects and patients with coarctation, between those with naCoA and reCoA, and between hypertensive and normotensive patients were evaluated using unpaired Student’s t tests assuming unequal variance between samples and Fisher exact χ 2 tests. Correlations between LVMi and other baseline exercise measurements were calculated using linear Pearson correlation. Exercise parameters, including LV lateral s′ and e′, septal s′ and e′, and IVA, were plotted against HR in regression models adjusted for repeated measures through an autoregressive covariance structure. All exercise parameters were modeled in a linear manner, while IVA was modeled as an exponential function; these models were selected because they provided the best fit to the data. In addition, IVA was plotted against HR to construct force-frequency curves for each individual. Slope of change of times was compared between control subjects and patients with CoA. Intraclass correlation coefficients were calculated for interobserver and intraobserver variability. A confidence level of P < .05 was considered statistically significant. Statistical analysis was performed with a commercially available package, SAS version 9.3 (SAS Institute Inc, Cary, NC).
Results
Clinical Data
Baseline characteristics of the two groups are summarized in Table 1 . The mean age was 17.8 ± 4.9 years; there was a male predominance of 87%. The mean age at initial stent implantation was 12.4 ± 3.5 years. The median follow-up duration (time from initial stent implantation to date of study) was 5.9 years (IQR, 1.6–7.0 years). After the initial stent implantation, five patients (17%) required additional interventions: balloon dilatation of the stent ( n = 3) and additional stent implantation ( n = 2). The median time from the latest intervention to the study was 2.7 years (IQR, 1.2–6.3 years).
| Variable | Control group | Coarctation group | P | reCoA group | naCoA group | P |
|---|---|---|---|---|---|---|
| n = 30 | n = 30 | n = 11 | n = 19 | |||
| Height (cm) | 167.8 ± 13.0 | 167.4 ± 14.0 | .91 | 170.1 ± 13.0 | 165.8 ± 14.7 | .42 |
| Weight (kg) | 61.6 ± 15.9 | 68.9 ± 17.1 | .09 | 72.6 ± 16.3 | 66.7 ± 17.6 | .37 |
| BSA (m 2 ) | 1.69 ± 0.28 | 1.78 ± 0.29 | .20 | 1.85 ± 0.27 | 1.74 ± 0.30 | .35 |
| Gender | ||||||
| Female | 4 (13%) | 4 (13%) | 1 (9%) | 3 (16%) | ||
| Male | 26 (87%) | 26 (87%) | 10 (91%) | 16 (84%) | ||
| Age at study (y) | 17.7 ± 5.3 | 17.8 ± 4.9 | .93 | 18.5 ± 5.1 | 17.4 ± 5.1 | .60 |
| Age at stent implantation (y) | 12.4 ± 3.5 | 13.9 ± 2.9 | 11.5 ± 3.5 | .05 | ||
| Duration of follow-up (y) | 5.9 (1.6–7.0) | 1.7 (0.4–7.2) | 6.0 (4.1–7.0) | .43 | ||
| Resting HR (beats/min) | 65 ± 11 | 65 ± 10 | .89 | 66 ± 11 | 64 ± 10 | .64 |
| Resting RA SBP (mm Hg) | 109 ± 10 | 119 ± 15 | .01 | 120 ± 18 | 117 ± 13 | .72 |
| Resting arm-leg SBP gradient (mm Hg) | 1 (0–9) | 0 (0–1) | 6 (0–12) | .09 | ||
| Resting arm-leg SBP gradient ≥ 20 mm Hg | 4 (13%) | 1 (9%) | 3 (16%) | |||
| 24-hr mean ABPM SBP (mm Hg) | 127 ± 12 | 127 ± 14 | 126 ± 11 | .81 | ||
| MASBP > 95th percentile | 13 (43%) | 5 (45%) | 8 (42%) | |||
| Hypertensive load (%) | 39 ± 27 | 41 ± 33 | 38 ± 25 | .82 | ||
| BP load > 40 mm Hg | 13 (43%) | 5 (45%) | 8 (42%) | |||
| Associated CHD | ||||||
| Bicuspid aortic valve | 14 (47%) | 5 (45%) | 9 (47%) | |||
| Others ∗ | 2 (7%) | 2 (18%) | 0 (0%) | |||
| Medications † | 3 (10%) | 1 (9%) | 2 (11%) |
∗ Subvalvar pulmonary stenosis and ventricular septal defect, status post ventricular septal defect repair (previous pulmonary artery banding), parachute mitral valve (mitral inflow mean gradient 2 mm Hg).
Two patients had hemodynamically insignificant intracardiac abnormalities (one patient had a repaired ventricular septal defect with no residual lesion, and the other had a parachute mitral valve with no mitral stenosis or regurgitation) considered not to influence the study, along with 14 patients (47%) with bicuspid aortic valves that were not stenotic or regurgitant.
reCoA versus naCoA
Eleven patients (37%) had reCoA and underwent initial surgical repair at a median age of 1 month (range, 10 days to 6.7 years). Repair was by subclavian flap in six (55%), end-to-end anastomosis in two (18%), and patch aortoplasty in three (27%). Because of residual arch obstruction, these patients underwent stent implantation at a mean age of 13.9 ± 2.9 years. The 19 patients (63%) with naCoA underwent stent implantation, at a mean age of 11.5 ± 3.5 years. There were no significant differences in demographics or clinical characteristics between those with naCoA and those with reCoA. There was, however, a trend toward younger age at stent implantation in the naCoA group.
Echocardiographic Data
Resting Echocardiographic Parameters
Resting echocardiographic parameters are presented in Tables 2 and 3 . Average resting gradient (corrected) across the aortic stent was 24 ± 11 mm Hg. Sixteen patients (53%) had resting gradients > 20 mm Hg (mean, 33 ± 7 mm Hg), with only three having resting arm-to-leg SBP gradients > 20 mm Hg (range, 22–30 mm Hg).
| Variable | Control group | Coarctation group | P | reCoA group | naCoA group | P |
|---|---|---|---|---|---|---|
| n = 30 | n = 30 | n = 11 | n = 19 | |||
| LV FS (%) | 37 ± 5 | 42 ± 6 | .002 | 39 ± 3 | 43 ± 7 | .09 |
| LVEF (%) | 62 ± 5 | 64 ± 7 | .12 | 63 ± 7 | 65 ± 7 | .29 |
| VCFC (circ/sec) | 1.15 ± 0.14 | 1.26 ± 0.20 | .02 | 1.18 ± 0.18 | 1.31 ± 0.20 | .08 |
| LVPWd (cm) | 0.69 ± 0.13 | 0.81 ± 0.15 | .004 | 0.84 ± 0.09 | 0.80 ± 0.17 | .37 |
| LVPWd Z score | −0.15 (−1.08 to 0.67) | 0.55 (−0.10 to 1.30) | .004 | 0.55 (0.28 to 1.23) | 0.50 (−0.43 to 1.3) | .43 |
| IVSd (cm) | 0.76 ± 0.15 | 0.87 ± 0.21 | .03 | 0.87 ± 0.24 | 0.86 ± 0.19 | .92 |
| IVSd Z score | 0.20 (−0.8 to 1.20) | 1 (0.15 to 2.03) | .04 | 0.80 (−0.45 to 1.95) | 1.17 (0.20 to 1.96) | .94 |
| LVEDd (cm) | 4.87 ± 0.46 | 5.31 ± 0.62 | .004 | 5.49 ± 0.75 | 5.20 ± 0.53 | .29 |
| LVEDd Z score | 0.20 (−0.60 to 0.68) | 1 (0.13 to 1.95) | .02 | 1.40 (0.63 to 2.05) | 0.80 (−0.175 to 1.69) | .56 |
| RWT | 0.29 ± 0.05 | 0.31 ± 0.05 | .17 | 0.31 ± 0.05 | 0.31 ± 0.05 | .77 |
| LVM (g) | 118.1 ± 34.4 | 192.8 ± 73.9 | <.001 | 208.6 ± 71.9 | 183.9 ± 75.6 | .40 |
| LVMi (g/m 2 ) | 68.9 ± 13.9 | 105.3 ± 31.4 | <.001 | 110.1 ± 30.9 | 102.6 ± 32.2 | .55 |
| LVM e score | −1.17 (−2.03 to −0.38) | 0.46 (−0.35 to 1.40) | <.001 | 0.15 (−0.29 to 1.59) | 0.49 (−0.31 to 1.35) | .65 |
| MV E (cm/sc) | 99 ± 17 | 118 ± 22 | <.001 | 116 ± 26 | 119 ± 21 | .72 |
| MV A (cm/sec) | 41 ± 11 | 57 ± 15 | <.001 | 57 ± 18 | 57 ± 14 | .99 |
| MV E/A ratio | 2.61 ± 0.79 | 2.18 ± 0.65 | .03 | 2.15 ± 0.66 | 2.21 ± 0.67 | .81 |
| DT (msec) | 156 ± 21 | 185 ± 60 | .02 | 179 ± 45 | 189 ± 69 | .68 |
| IVRT (msec) | 76 ± 9 | 77 ± 8 | .84 | 80 ± 9 | 75 ± 7 | .11 |
| PV S (cm/sec) | 40 ± 12 | 50 ± 13 | .003 | 54 ± 15 | 48 ± 12 | .23 |
| PV D (cm/sec) | 61 ± 13 | 62 ± 13 | .74 | 62 ± 13 | 62 ± 13 | .88 |
| PV Ar (cm/sec) | 23 ± 19 | 28 ± 8 | .35 | 27 ± 5 | 28 ± 9 | .82 |
| PV Ar duration (msec) | 107 ± 41 | 115 ± 23 | .49 | 115 ± 19 | 115 ± 26 | .99 |
| PV ArD − MV AD (msec) | −10 (−21 to −2) | −17 (−46 to 6) | .53 | −9 (−56 to 5) | −17 (−45 to 6.5) | .70 |
| PV Ar/MV A | 0.60 ± 0.51 | 0.50 ± 0.16 | .48 | 0.49 ± 0.16 | 0.51 ± 0.17 | .83 |
| Resting stent gradient (mm Hg) ∗ | 24 ± 11 | 27 ± 10 | 23 ± 11 | .38 |
∗ Gradient across aortic stent at rest, corrected for flow velocity proximal to stent.
| Variable | Control group | Coarctation group | P | reCoA group | naCoA group | P |
|---|---|---|---|---|---|---|
| n = 30 | n = 30 | n = 11 | n = 19 | |||
| LV lateral e′ (cm/sec) | 18.1 ± 2.3 | 17.6 ± 3.2 | .45 | 18.1 ± 3.9 | 17.3 ± 2.8 | .54 |
| LV lateral a′ (cm/sec) | 6.1 ± 1.8 | 5.8 ± 1.9 | .59 | 5.4 ± 1.1 | 6.1 ± 2.2 | .22 |
| LV lateral s′ (cm/sec) | 11.5 ± 2.6 | 11.1 ± 2.8 | .49 | 9.9 ± 1.7 | 11.8 ± 3.1 | .04 |
| IVS e′ (cm/sec) | 15.2 ± 2.1 | 14.5 ± 2.8 | .27 | 14.7 ± 3.1 | 14.4 ± 2.7 | .78 |
| IVS a′ (cm/sec) | 6.2 ± 1.4 | 7.5 ± 1.5 | <.001 | 7.9 ± 1.5 | 7.3 ± 1.5 | .32 |
| IVS s′ (cm/sec) | 8.7 ± 0.9 | 9.0 ± 1.7 | .41 | 8.9 ± 1.3 | 9.1 ± 1.9 | .70 |
| LV lateral e′/a′ ratio | 3.17 ± 0.83 | 3.23 ± 1.04 | .80 | 3.56 ± 1.29 | 3.04 ± 0.84 | .25 |
| LV e/e′ ratio | 5.56 ± 0.98 | 6.92 ± 1.87 | <.001 | 6.64 ± 2.14 | 7.09 ± 1.73 | .56 |
| IVS e′/a′ ratio | 2.56 ± 0.59 | 1.99 ± 0.49 | <.001 | 1.95 ± 0.66 | 2.01 ± 0.38 | .79 |
| IVS e/e′ ratio | 6.69 ± 1.44 | 8.69 ± 2.73 | <.001 | 8.20 ± 2.65 | 8.97 ± 2.80 | .46 |
| LV IVA (m/sec 2 ) | 1.02 ± 0.31 | 0.85 ± 0.30 | .04 | 0.84 ± 0.29 | 0.86 ± 0.31 | .89 |
| Global LV systolic longitudinal strain (%) | −19.5 ± 1.8 | −19.2 ± 2.2 | .59 | −18.3 ± 2.1 | −19.8 ± 2.2 | .10 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
