Usefulness of Doppler echocardiography to determine the timing of surgery for supravalvar aortic stenosis

Abstract

S aortic stenosis (SVAS) may occur as part of Williams’ syndrome, with autosomal dominant inheritance or sporadically.1–3 Previous studies have shown that SVAS tends to progress with time, so serial evaluation is important.4–6 Operative intervention for SVAS has been recommended for symptoms or for a catheter-measured gradient of .50 mm Hg.2,3,7,8 Although previous studies have shown the value of 2-dimensional and Doppler imaging in the diagnosis of SVAS,9–14 the value of echocardiography for predicting the catheter-measured gradient and assessing the need for surgery has not been evaluated. This study assesses the value of echocardiography and Doppler imaging in predicting catheter-measured pressure gradients and determining the need for surgery for SVAS. • • • The Primary Children’s Medical Center pediatric cardiology database was searched to identify patients with the diagnosis of SVAS who underwent echocardiography within 2.5 months of catheterization. Demographic data collected included patient weight, body surface area, gender, age at catheterization, and surgical history. Complete 2-dimensional, M-mode, and Doppler echocardiograms were retrospectively reviewed. Measurements were made off-line from digitized images selected from videotape. Each measurement was made in triplicate and averaged. Two-dimensional measurements were made from inner edge to inner edge. The narrowest diameter of the ascending aorta and the aortic annulus diameter were measured, allowing calculation of a ratio (supravalvar aorta/annulus) that was always ,1. In each case, color Doppler was used to align the continuous-wave Doppler beam as close to parallel as possible with the jet in the ascending aorta, but no correction was made for angle of incidence. The peak instantaneous and mean gradients were obtained from Doppler tracings (modified Bernoulli equation). The presence of aortic regurgitation and the presence of head and neck vessel or branch pulmonary artery stenoses were noted. Left ventricular shortening fraction, wall thickness, and left ventricular mass (American Society of Echocardiography convention) were measured from the M-mode tracing.16 A cardiologist blinded to the results of the echocardiograms reviewed catheterization data and angiograms. In each case, the peak-to-peak pressure gradient, the type of SVAS (discrete, diffuse, or membranous), the ratio of the ascending aortic narrowing to the aortic annulus, the presence of stenoses of other vessels (head and neck vessels, coronaries, branch pulmonary arteries, renal arteries), and the presence of aortic regurgitation were noted. Linear regression analysis was used to assess the correlation between Doppler and catheter gradients, and between other variables and Doppler or catheter gradients. The influence of variables on the difference between Doppler and catheter gradients was assessed using either linear regression or unpaired t tests. Agreement between Doppler and catheter gradients was assessed using the method of Bland and Altman, which involves plotting the difference between the Doppler and catheter gradients against the average of the same 2 gradients for each case.16 The mean difference between the 2 measurements gives an indication of the bias, and the limits of agreement are defined by the mean difference 6 2 SDs (95% confidence intervals). When there is good agreement between 2 measurements, there is an expected small mean difference and narrow limits of agreement. Because of the poor agreement between Doppler and catheter gradients, we then assessed whether Doppler might be clinically useful in predicting a catheter-measured peak gradient of .50 mm Hg, the currently published criterion for surgery for SVAS.2,3,7,8 The relation between a Doppler peak instantaneous gradient of .85 mm Hg and a catheter peak-to-peak gradient of .50 mm Hg was assessed using Fisher’s exact test. Sensitivity, specificity, positive and negative predictive values, and accuracy of a Doppler peak instantaneous gradient of .85 mm Hg for predicting a catheter-measured peak-to-peak gradient of .50 mm Hg were also calculated. A p value of ,0.05 was considered statistically significant. A total of 26 paired echocardiograms and catheterizations from 18 patients (12 males, 6 females) with SVAS were identified. When .1 echocardiographic catherterization pair was used from the same patient, the time interval ranged from 6.3 to 108 months (median 22.5). Age at catheterization ranged from 1 month to 18 years (median 2.5 years). From the aortic angiograms, the SVAS was discrete or localized in 20 cases, and diffuse in the other 6. Membranous obstruction was not seen in this series. Sedation was required for 10 echocardiograms. The other patients were cooperative, and studied in a quiet, resting state. The Doppler tracing used for determination of the peak instantaneous and mean gradients was obtained using blind continuous-wave Doppler in 11 cases and by 2-dimensional directed continuous-wave Doppler in From the Department of Pediatrics, Primary Children’s Medical Center and the University of Utah, Salt Lake City, Utah. Dr. Tani’s address is: Primary Children’s Medical Center, 100 North Medical Drive, Salt Lake City, Utah 84113. Manuscript received and accepted December 2, 1999; revised manuscript received and accepted January 18, 2000.