Aortic root dilation in bicuspid aortic valve disease: impact on dissection risk and evaluation of guideline-based surgical thresholds — a retrospective cohort study

Figures

• 

  Figure 1.. Growth rates of the ascending aorta and aortic root in patients with root and non-root dilation. (A) Violin plots illustrating the estimated ascending aortic growth rate and aortic root growth rate in patients with root dilation (orange) and non-root dilation (green). The growth rates are expressed in millimeters per year (mm/year). The accompanying table summarizes the mean ± SD and P values of the T-test for comparisons between the two groups. (B) Scatter plot with regression lines showing the relationship between age and the estimated ascending aortic growth rate in patients with root dilation (orange) and non-root dilation (green). The regression curves are fitted using GAMs, with shaded regions representing the 95% confidence interval (CI). (C) Scatter plot with regression lines depicting the association between age and the estimated aortic root growth rate for the two patient groups. The regression curves are fitted using GAM, with shaded regions representing the 95% CI. CI, confidence interval; SD, standard deviation.

• 

  Figure 2.. Cumulative incidence of AAD during follow-up. (A) Cumulative incidence curve of AAD for the overall cohort over a 10-year follow-up period. The number at risk for each group is displayed below the curve. (B) Cumulative incidence curves of AAD stratified by dilation type (root dilation: orange; non-root dilation: green). The comparison between groups was performed using the log-rank test, which showed a significantly higher incidence of AAD in patients with root dilation (P value = 0.02). The number at risk for each group is displayed below the curve. AAD, Stanford type A aortic dissection.

• 

  Figure 3.. HR for AAD and the first derivative of HR with aortic diameter. (A) The relationship between the ascending aortic diameter (mm) and the HR for AAD. (B) The first derivative of the HR curve from panel (A), indicating the rate of change in HR (dHR/dx) as the ascending aortic diameter increases. (C) The relationship between the aortic root diameter (mm) and the HR for AAD. (D) The first derivative of the HR curve from panel (C), showing the rate of change in HR (dHR/dx) as the aortic root diameter increases. AAD, Stanford type A aortic dissection; HR, hazard ratio.

• 

  Figure 4.. DCA and ROC curves for aortic dissection prediction based on aortic root and ascending aorta diameter thresholds. (A–B) DCA comparing the net clinical benefit of applying predefined surgical thresholds versus treating all or treating none. Panel A shows the DCA for an aortic root diameter ≥45 mm, and Panel B shows the DCA for an ascending aortic diameter ≥50 mm. Vertical dashed lines indicate the threshold probabilities corresponding to these diameter cutoffs. (C–D) ROC curves evaluating the discriminative performance of different aortic diameter thresholds for AAD. Panel C shows AUCs for aortic root diameter cutoffs of 45 mm (AUC = 0.738), 50 mm (AUC = 0.572), and 55 mm (AUC = 0.500). Panel D shows AUCs for ascending aortic diameter cutoffs of 45 mm (AUC = 0.550), 50 mm (AUC = 0.627), and 55 mm (AUC = 0.500). AAD, Stanford type A aortic dissection; AUC, area under the curve; DCA, decision curve analysis; ROC, receiver operating characteristic.

• 

  Figure 5.. Assessment of guideline-based models for surgical timing: (A) Summary of surgical intervention thresholds for BAV-related aortic dilation in different guideline models and the hypothetical model. The figure summarizes the surgical intervention thresholds for BAV-related aortic dilation recommended by the 2014 ESC Guidelines, 2022 ACC/AHA Guidelines, 2024 ESC Guidelines, and the hypothetical model proposed in this study. (B) Comparative ROC curves for the four guideline-based models in predicting AAD risk. The hypothetical model proposed in this study achieves the highest predictive performance (AUC = 0.752), followed by the 2024 ESC model (AUC = 0.731), the 2022 ACC/AHA model (AUC = 0.721), and the 2014 ESC model (AUC = 0.712). AAD, Stanford type A aortic dissection; AUC, area under the curve; BAV, bicuspid aortic valve; ROC, receiver operating characteristic.

• 

  Figure 6.. Surgical treatment patterns and competing risk analysis of AAD in patients with varying aortic diameters. (A) Aortic surgery rates stratified by aortic size (45–50, 51–55, and >55 mm) and measurement location (ascending aorta vs. aortic root). Surgery was more frequently performed for root diameters between 45 and 50 mm compared to ascending diameters in the same range (65.3% vs. 47.9%, P value < 0.001), whereas differences were not significant in larger size categories. (B) Cumulative incidence curves of AAD derived from a competing risk model using the Gray–Fine method. The predicted curve (red) represents the hypothetical cumulative incidence of AAD if all patients were assumed not to undergo surgery. The observed curve (green) shows the actual AAD incidence under real-world surgical practice. The divergence between the two curves highlights the potential protective effect of surgery on AAD occurrence. (C) Subgroup analysis of cumulative AAD incidence based on aortic root diameter: <45 mm (left) and ≥45 mm (right). Predicted incidence without surgery (red) was consistently higher than observed incidence with surgery (green), especially in patients with root diameters ≥45 mm,…