A clinical and echocardiographic model to identify LVOT-related overestimation of aortic stenosis severity in transcatheter aortic valve replacement candidates

Figures

• 

  Figure 1. CONSORT diagram of study population. Following inclusion and exclusion criteria, a total of 1367 individuals were included in our analysis.

• 

  Figure 2. Bland–Altman plots comparing CT- and echocardiographic measurements of LVOT area and resulting AVA. (A) The left panel displays the Bland–Altman plot comparing LVOT area measured by echocardiography and by CT. The plot reveals a systematic underestimation of LVOT area by echocardiography compared to CT derived direct measurement, with increasing bias at larger mean values. (B) The right panel shows the Bland–Altman plot comparing echo-derived AVA and hybrid AVA. A proportional trend is observed, with decreasing agreement at higher AVA values. The red line indicates the regression of the differences against the means with 95% confidence intervals; dashed lines represent the bias, and the limits of agreement (bias ± 1.96 SD).

• 

  Figure 3. Correlation between hybrid and predicted AVA in the derivation and validation cohorts. Scatter plot and regression line with 95% confidence interval for AVA values predicted by the parsimonious model using only echocardiographic and patient characteristics vs. hybrid AVA values. (A) derivation cohort on which the parsimonious model was developed. (B) validation cohort which was not used in any way to derive the parsimonious model.

• 

  Figure 4. Probability of severe and non-severe AS based on hybrid and predicted AVA across the range of echo-derived AVA values, along with the distribution of echo-derived AVA values within the study population. Yellow line shows the probability of non-severe AS by hybrid AVA and lilac line corresponds to the probability of non-severe AS by predicted AVA across the echo AVA ranges. The grey histogram in the background shows the distribution of echo AVA values in our population of 1367 individuals .

• 

  Figure 5. Kaplan–Meier survival analysis in the derivation cohort with comparison between groups stratified by hybrid AVA. Red line shows patients reclassified to moderate AS (hybrid AVA >1.2 cm2), green line indicates patients with indeterminate AS (hybrid AVA 1.0–1.2 cm2), and blue line presents patients with true severe AS (hybrid AVA <1.0 cm2).

• 

  Figure 6. Predicted AVA values for different echocardiographic LVOT diameter and Doppler index values in females and males. (A) Represents predicted AVA values for female patients. (B) Predicted AVA values for male patients. The red contour line indicates the 1.0 cm2 threshold of echo-derived AVA, while the black lines represent 1.0 and 1.2 cm2 levels of the predicted AVA.

• 

  Figure 7. Application of the clinical chart illustrated by two representative cases. Case 1: A 77-year-old female with low-flow, low-gradient AS (AVA 0.46 cm2). Based on an LVOT diameter of 19 mm and a DI of 0.19, the clinical chart predicted a hybrid AVA of 0.7–0.8 cm2. The value derived from direct CT LVOT measurement was 0.61 cm2, indicating true severe AS. Case 2: A 90-year-old female with normal-flow, low-gradient AS (AVA 0.96 cm2). With an LVOT diameter of 22 mm and a DI of 0.26, the chart estimated a hybrid AVA of 1.2–1.3 cm2. The CT-derived value was 1.27 cm2, suggesting moderate AS .