DETERMINATION OF AORTIC-VALVE AREA BY 2-DIMENSIONAL AND DOPPLER ECHOCARDIOGRAPHY IN PATIENTS WITH NORMAL AND STENOTIC BIOPROSTHETIC VALVES

To assess the feasibility and accuracy of determining bioprosthetic aortic valve area from two-dimensional and Doppler echocardiographic measurements, three partially overlapping groups were selected from 55 patients with such bioprosthetic valves and adequate Doppler studies. These were Group 1, 37 patients with recent aortic valve replacement surgery and no clinical or echocardiographic evidence of valve dysfunction; Group 2, 12 patients with prosthetic valve stenosis documented by cardiac catheterization; and Group 3, 22 patients with both Doppler and catheterization studies in whom noninvasive and invasive determinations of aortic valve area could be directly compared. Left ventricular outflow tract diameter was measured from two-dimensional still frame images. Flow velocity proximal to the aortic valve, transvalvular velocity and acceleration time were determined from pulsed and continuous wave Doppler spectra. Aortic valve gradient was calculated with the modified Bernoulli equation and valve area by the continuity equation. In the 37 patients with a normally functioning valve, the calculated mean gradient ranged from 5 to 25 mm Hg (average 13.6 ± 5.2) and valve area from 1.0 to 2.3 cm2 (mean 1.6 ± 0.31). Linear regression analysis of prosthetic aortic valve area determined by Doppler imaging and cardiac catheterization demonstrated a high correlation (r = 0.93) between the two techniques. Comparison of the patients with and without prosthetic valve stenosis revealed statistically significant differences in mean gradient (42.8 ± 12.3 versus 13.6 ± 5.2 mm Hg; p = 0.0001), acceleration time (116 ± 15 versus 80 ± 13 ms; p = 0.0001) and valve area by the continuity equation (0.80 ± 0.16 versus 1.6 ± 0.31 cm2; p = 0.0001). Individual ultrasound variables were assessed for their utility in recognizing bioprosthetic valve stenosis. When diagnostic criteria were selected to maintain absolute specificity, either an abnormally high mean gradient or a noninvasively determined valve area <1 cm2 identified 92% of patients with a stenotic valve. Marginally lower sensitivity was associated with an elevated peak gradient or a prolonged acceleration time; each identified 75% of the patients with valve stenosis. However, measurement of the ratio of left ventricular outflow tract to transvalvular velocity time integrals further improved diagnostic accuracy. All 37 patients with a normal prosthetic valve had a ratio >0.35, whereas each of the 12 patients with valve stenosis had a lower ratio. In conclusion, assessment of prosthetic aortic valve area by cardiac ultrasound is highly accurate and can be useful in the diagnosis of bioprosthetic aortic valve stenosis. © 1990.