Echocardiographic Versus Invasive Aortic Valve Gradients in Different Clinical Scenarios

Abstract

The role of echocardiography in deriving transvalvular mean gradients from transaortic velocities in aortic stenosis (AS) and in structural valve degeneration (SVD) is well established. However, reports following surgical aortic valve replacement, post-transcatheter aortic valve replacement (TAVR), and valve-in-valve-TAVR (ViV-TAVR) have cautioned against the use of echocardiography-derived mean gradients to assess normal functioning bioprosthesis due to discrepancy compared with invasive measures in a phenomenon called discordance. In a multicenter study, intraprocedural echocardiographic and invasive mean gradients in AS, SVD, post-native TAVR, and post-ViV-TAVR were compared, when obtained concomitantly, and discharge echocardiographic gradients were recorded. Absolute discordance (intraprocedural echocardiographic - invasive mean gradient) and percent discordance (intraprocedural echocardiographic - invasive mean gradient/echocardiographic mean gradient) were calculated. Multivariable regression analysis was performed to determine variables independently associated with elevated postprocedure invasive gradients ≥20mm Hg, absolute discordance >10mm Hg, and discharge echocardiographic mean gradient ≥20mm Hg. A total of 5,027 patients were included in the registry: 4,725 native TAVR and 302 ViV-TAVR. Intraprocedural concomitant echocardiographic and invasive mean gradients were obtained pre-TAVR in AS (n=2,418), pre-ViV-TAVR in SVD (n=101), in post-ViV-TAVR (n= 77), and in post-TAVR (n= 823). Echocardiographic and invasive mean gradients demonstrated strong correlation (r=0.69) and agreement (bias, 0.11; 95% CI, -0.4-0.62) in AS, moderate correlation (r=0.56) and agreement (bias, 1.08; 95% CI, -2.53 to 4.59) in SVD, moderate correlation (r=0.61) and weak agreement (bias, 6.47; 95% CI, 5.08-7.85) post-ViV-TAVR, and weak correlation (r=0.18) and agreement (bias, 3.41; 95% CI, 3.16-3.65) post-TAVR. Absolute discordance occurs primarily in ViV-TVR and is not explained by sinotubular junction size and increases with increasing echocardiographic mean gradient. Percent discordance in AS and SVD (1.3% and 4%, respectively) was lower compared with post-TAVR/ViV-TAVR (66.7% and 100%, respectively). Compared with self-expanding valves, balloon expanding valves were independently associated with elevated discharge echocardiographic but lower invasive mean gradient (odds ratio=3.411, 95% CI, 1.482-7.852, P=.004; vs odds ratio=0.308, 95% CI, 0.130-0.731, P=.008, respectively). Post-TAVR/ViV-TAVR, echocardiography is discordant from invasive mean gradients, and absolute discordance increases with increasing echocardiographic mean gradient and is not explained by sinotubular junction size. Percent discordance is significantly higher post-TAVR/ViV-TAVR than in AS and SVD. Post-TAVR/ViV-TAVR, poor correlation and wide limits of agreement suggest echocardiographic and invasive mean gradients may not be used interchangeably and a high residual echocardiographic mean gradient should be confirmed invasively before considering any additional procedure to "correct" the gradient. Transcatheter aortic valve replacement valve types have variable impact on echocardiographic and invasive mean gradients.