Main

When performing an echocardiography in children with coarctation of the aorta, it is commonly observed that the pulsatility of the poststenotic aorta. i.e. increase in aortic diameter due to the pulse wave pressure after cardiac systole, is impaired in these patients (Fig. 1). This is above all due to the loss of the pulse wave pressure at the site of vessel obstruction.

Figure 1
figure 1

Left panel, M-mode echocardiographic image of normal pulsatility of the abdominal aorta in a 5-y-old boy without heart disease. Right panel, typically decreased pulsatility during the cardiac cycle in the abdominal aorta of a 5-y-old boy with severe coarctation.

Full size image

Doppler echocardiography as the mainstay of the diagnostic work-up in patients with coarctation has been shown to be associated with some pitfalls in the appreciation of severity(1) and even the presence of coarctation(2) in pediatric populations. In addition, although in infants usually excellent two-dimensional echocardiographic visualization of the site of obstruction is possible, this is not the case for some of the older children. Moreover, the clinical hallmarks of the disease may not be reliable in all cases and systemic hypertension may be lacking or the femoral pulses may be palpable despite the presence of severe coarctation.

Quantitative description of impaired pulsatility of the post-stenotic aorta and comparison of this pulsatile characteristic with normal children has not been previously investigated. In our study the hypothesis was tested, that altered mechanical behavior of the abdominal aorta due to the loss of the pulse wave pressure, when assessed quantitatively, produced consistent and reproducible differences between affected patients and normals.

PATIENTS AND METHODS

Patients. Beginning in January 1996 and ending in October 1997, all the children diagnosed in our institution as having severe coarctation of the aorta were prospectively evaluated for the purpose of our study. A total of 27 children were included; 19 children (15 male, 4 female, median age 5.1 y, range 0.2-12.1) had unoperated coarctation, the other 8 patients (5 male, 3 female, median age 9.3 y, range 2.8-13.5) had recoarctation after initial surgical treatment at a mean of 5.9 y before the current evaluation. A group of 23 healthy children, matched with the coarctation patients for age, sex distribution (16 male, 7 female) and weight, were studied according to the same protocol. Structural heart disease was excluded in all of these normal children on clinical grounds and with a complete echocardiographic evaluation.

Severe coarctation was defined as the presence of a blood pressure gradient of >20 mm Hg between upper and lower limbs(3) or a lower gradient with arterial hypertension (right arm blood pressure >95 percentile for age) and/or a corrected Doppler gradient in the descending aorta of >20 mm Hg. Patients were included if they presented with severe native coarctation beyond the neonatal age (>1 mo of age) or with severe recurrent coarctation after earlier surgical treatment, all these patients were studied during the routine diagnostic work-up before therapeutic intervention. For inclusion in the study, an angiographic documentation of the site and extent of the aortic narrowing was required. Excluded were all children, where echocardiographic evaluation of the abdominal aorta resulted in poor image quality, inadequate for proper measurements and calculations (n = 2). Children were not included if they had associated congenital heart disease of any kind, except for the presence of no more than mild valvar aortic stenosis (echo Doppler gradient <20 mm Hg) or only trivial valvar regurgitation. Also excluded were patients with impaired left ventricular function (shortening fraction in the M-mode echocardiogram <30%). Finally, all patients with vasoactive drug therapy of any kind were not considered for evaluation. Normal controls were recruited among children seen at the pediatric cardiology outpatient clinic for evaluation of a cardiac murmur and after echocardiographic documentation of a normal heart.

Oral informed consent was obtained from patients and/or parents and the study had been approved by the institution's committee on research.

Evaluation of coarctation. All patients had a complete M-mode, two-dimensional, and color-Doppler echocardiographic study of cardiac structures including assessment of aortic arch anatomy. Severity of coarctation was assessed from the suprasternal notch with two-dimensional imaging of the descending aorta and with Doppler measurement of blood flow velocity in the descending aorta, correcting the Doppler gradient for the blood flow velocity in the prestenotic aorta according to the original Bernoulli formula [4 (v12 - v22) where v1 is the velocity in the descending and v2 the velocity in the prestenotic aorta]. The maximum Doppler gradient obtained in a given patient was used as echocardiographic estimation of severity of coarctation. Angiography (i.v. digitized subtraction angiography in 18 patients; angio-magnetic resonance imaging in nine children) was used as the reference for the severity of coarctation by calculating the relative aortic narrowing (minimum aortic diameter as percentage of the aortic diameter measured between the left carotid and the left subclavian artery).

Blood pressure measurements were performed with the oscillometric method (Dinamap, Critikon) using an appropriate sized cuff(4). The mean of three single measurements with the child quiet and supine was taken on all four limbs. Normal values for blood pressures in children were taken from the percentile curves established by the Task force on blood pressure control in children(5). Clinical estimation of severity of coarctation in the individual patient was made by comparing the peak systolic right arm with the peak systolic right leg blood pressure. Variability of blood pressure measurements was a maximum of 9% of the mean in any individual child (variability coefficient of 0.07).

Echocardiographic measurements of abdominal aorta. Echocardiographic examination of the aorta was performed in a similar way as previously validated (in adults) for the ascending aorta(6) with the unsedated patients supine. An Acuson 128 XP 10 machine (Acuson, Mountain View, CA) with a 5 MHz transducer was used in all instances. A simultaneous one-lead ECG recording was displayed on the monitor at a speed of 50 mm/s. The sector array echo transducer was kept longitudinally in the epigastric space to obtain an optimal two-dimensional image of the aorta. The M-mode beam was kept perpendicular to the posterior aortic wall, proximal to the superior mesenteric artery. Then the image was magnified using the resolution box function. Depth and gain settings were individually optimized to best contrast the vessel walls from the lumen.

The inner diameter of the aortic lumen (measured between the trailing edge of the anterior aortic wall and the leading edge of the posterior aortic wall) was measured in diastole (Dd) as well as in systole (Ds). The time interval was measured between Dd and Ds as the tu (expressed as ms) (Fig. 2). The RR interval was measured on the simultaneous ECG. The computer software of the ultrasound equipment indicated a calibration for time scale and for the real distance on each image displayed.

Figure 2
figure 2

Schematic drawing of the measurements taken in the M-mode image of the abdominal aorta (see figure legends for explanation of abbreviations).

Full size image

Of the measurements taken, the following age- and weight-independent indexes were calculated: the percent increase of the abdominal aortic inner diameter (expressed as % of the Dd), and the increase in aortic wall inner diameter, indexed to the upstroke time tu (D/tu expressed as %/s). For off-line measurements and calculations, the frozen M-mode image was printed on thermal paper using a Mitsubishi K70S video copy processor. All values used for calculations were taken as the mean of three single measurements of three consecutive cardiac cycles. All the echocardiographic examinations and calculations were performed with the prints coded and the observer blinded for the diagnosis.

For calculation of the intraobserver variability, measurements from 15 prints (three identical copies for 5 different patients and normals), blinded for the observer were compared.

Statistical evaluation. Values are shown as mean ± SD and as the median with range. Comparison between groups was made using the unpaired two-tailed t test or the Wilcoxon rank sum test. Variability of blood pressure measurements and aortic measurements was expressed as a percentage and by calculating the coefficient of variation (the SD of the duplicate measurements is calculated from the equation √Σd2/2n, where d is the difference between duplicates and n the number of pairs, its expression relative to the mean of the duplicate measurements is the coefficient of variation). For calculation of a correlation between either coarctation gradients or measured relative aortic narrowing with the pulsatility index, univariate linear regression was performed using a commercially available software (Systat 5.01). p < 0.05 was considered statistically significant.

RESULTS

During the study period, 27 pediatric patients with significant coarctation of the aorta could be evaluated according to the study protocol: 25 of these 27 patients had systemic hypertension with systolic blood pressure values above the 95th percentile for age. The blood pressure gradient varied between 10 and 80 mm Hg (median 40 mm Hg) in these children, the corrected echo-Doppler gradients varied between 26 and 64 mm Hg (median 42, p = NS). Continuous wave Doppler interrogation of the descending aorta in patients resulted in a typical diastolic gradient of the Doppler spectrum in 20/27 affected children. Femoral pulses were absent in 22/27 patients. Angiographic assessment of the severity of stenosis showed a discrete short segment stenosis in 24 patients and a longer stenotic aortic segment in the remaining 3 patients. The aorta showed a minimum residual lumen of between 17 and 40% (median 31, mean 32 ± 6%) of the prestenotic aortic diameter. A comparison of clinical data of the patient groups and normal controls is shown in Table 1.

Table 1 Comparison of clinical data of study groups
Full size table

Echocardiographic Measurements

Normal children. Median of minimum abdominal aortic diameter was 7.2 mm (range 5.2-12.2) and increased to a median of 10.4 mm (range 6.9-15.7) for the maximum aortic diameter. The mean increase of aortic diameter was 37 ± 8% of minimum diameter. Mean time of upstroke from minimum to maximum diameter was 185 ± 33 ms in these normals, or maximum aortic distension at 28 ± 4% of the cardiac cycle time. The increase in abdominal aortic diameter indexed to a unit of time showed a mean of 15.2 ± 3 mm/sec or 201 ± 33%/sec (detailed values shown in Table 2).

Table 2 Comparison of echocardiographic measurements and calculations of poststenotic aortic wall pulsatility in different groups
Full size table

Children with relevant aortic obstruction. A total of 19 children had echocardiographic measurements of the abdominal aorta before surgical treatment and 8 more patients with recurrent coarctation after initial surgical resection had abdominal aorta measurements performed before surgical reintervention (n = 4) or to balloon dilatation (n = 4). In patients with severe coarctation, mean minimum abdominal aortic diameter was no different from controls but the mean maximum diameter in patients was significantly smaller than in normal children (p < 0.005; detailed values shown in Table 2). The mean increase in the aortic diameter did not differ statistically between unoperated patients (median 14, range 5-23) and those with recoarctation (median 18, range 8-25; p = NS). These values differed significantly from those obtained in normal controls (Table 2). Analyzing the individual values it was observed, that there was not only a statistical difference, but that also there was no overlap of individual values between normals and patients with relevant aortic obstruction. No patient had a percent increase of the aortic diameter of more than 25% whereas none of the controls had a value below 27% (Fig. 3). The same finding was obtained regarding the percent increase of aortic diameter indexed to unit of time: patient groups with relevant aortic obstruction did not differ between each other with means of 51 ± 24%/s (median 50; range 18-108) in unoperated patients and of 68 ± 25%/s (median 69; range 30-106) in children with recurrent coarctation. Again these values differed significantly from those obtained in controls (p < 0.001) and showed no overlap of individual values between patients and normals as shown in Figure 4. The mean time of upstroke to maximum diameter was significantly longer in children with coarctation when compared with normals (p < 0.01) but these results showed a clear range of overlap of individual measurements between groups. The same was found for the upstroke time expressed as a fraction of the cardiac cycle time (detailed values given in Table 2). A significant association was found between the increasing severity of stenosis (as assessed with right arm - right leg blood pressure gradient) and the indexed percent increase of the aortic diameter (p < 0.01, r = -0.72) but there was no significance in the association between the anatomically measured relative aortic narrowing and the index of pulsatility.

Figure 3
figure 3

Comparison of the percent increase in abdominal aortic diameter in the study groups. Values are shown as median (circles) and range. Horizontal bar indicating the cut-off value between patients with relevant coarctation and normal children (Dd, minimum aortic diameter).

Full size image
Figure 4
figure 4

Comparison of indexed percent increase of abdominal aortic diameter in the different groups of children. Values are shown as median (circles) and range.

Full size image

Repetitive measurements were performed to assess the intraobserver variability of measurements and calculations. For the individual measurements, intraobserver variability was up to 6% for the diameter measurements and up to 12% for the measurements of time intervals. By taking all values as mean of three single measurements, the intraobserver variability decreased to 2% for diameter measurements and to 4% for measurements of time intervals. For calculations of indexes, this led to a variability of up to 6% (coefficient of variation of 0.055).

DISCUSSION

In children with coarctation, clinical findings such as systemic hypertension and absent femoral pulses together with two-dimensional and Doppler-echocardiographic evaluation allow for an adequate assessment of the presence and severity of coarctation in a majority of the children examined(7,8). Nevertheless, there are pediatric patients with significant aortic obstruction and absence of the typical clinical findings as it was again shown by our study population with 7 of 27 children with proven relevant stenosis having either palpable femoral pulses or absence of systemic hypertension. Moreover, echocardiography and Doppler analysis of coarctation in particular have been shown to be associated with sometimes severe pitfalls and not to be entirely reliable in the assessment of severity of coarctation in a certain number of cases studied(1,2,9). Thus, additional noninvasive tools to better estimate the severity of coarctation would be helpful in the diagnostic work-up of these children.

The main findings of our study were that quantitative description of impaired pulsatility of the abdominal (poststenotic) aorta allowed for a clear discrimination between patients with relevant stenosis and normal controls. Although a significant inverse relation between increasing Doppler gradient over the site of obstruction and decreasing pulsatility index was found, a similar association could not be established between increasing severity of aortic obstruction (as assessed using angiographic data) and the decrease in the pulsatility of the poststenotic aorta. Based on the data presented it may be assumed that patients with < 40% of residual aortic lumen exhibit an impaired pulsatility of the poststenotic aorta that allowed for a clear discrimination from normal children.

Echocardiographic studies in the descending aorta have been previously performed but they solely concentrated on altered Doppler flow profiles of the poststenotic aorta. It was shown that quantitative assessment of pulsed Doppler curves in the descending aorta at the level of the diaphragm resulted in consistent and reproducible findings that allowed for a discrimination between patients with coarctation and normals(10,11). These studies have never been extended to assess the alteration in pulsatility of the abdominal aorta that has the same cause as the alterations shown for the Doppler flow profiles, namely the partial loss of the pulse wave pressure at the site of obstruction. As the angle for Doppler interrogation of the thoracic and abdominal aorta is far from ideal, pulsatility analysis of the poststenotic aorta would be easier to obtain and reproduce as an additional noninvasive diagnostic tool.

In our study it was demonstrated that in addition to these previously performed Doppler flow profile analysis in the descending aorta, also the assessment of poststenotic aortic pulsatility, using four measures obtained with a standardized M-mode echocardiographic study, allowed for a significant differentiation between patients and normals. Moreover, it was possible to define two age and body-size independent indexes (the percent increase of the aortic diameter and that percentage indexed per unit of time), that allowed for a clear discrimination between patients with isolated coarctation and normal children with no overlap of individual values.

An ongoing extension of this study now has to test whether the method remains of value in the presence of associated cardiac defects. As shown for the altered pulsed Doppler flow patterns in the descending aorta(10), it must be assumed that an arterial run-off (ductus, collaterals) or left ventricular dysfunction might also influence the pattern of the mechanical behavior of the poststenotic aorta. Furthermore, it is currently prospectively tested, what characteristics of abdominal aortic pulsatility are obtained in patients with mild forms of aortic obstruction and in which clinical situations this additional diagnostic indexes might be of value in the decision making in children with coarctation.

Study Limitations

For the initial evaluation of this new echocardiographic application in children with coarctation, it was necessary to first describe a homogenous study group with a strictly defined pathological condition (isolated coarctation) and to exclude all patients with associated cardiac anomalies that clearly were found in a large proportion of pediatric patients with coarctation. Thus, conclusions on the results obtained currently apply to only a well-defined small group of coarctation patients. Furthermore, our data need to be verified in a larger population of affected children.