Ratio of vena contracta width to aortic valve area index for risk stratification of patients with mixed aortic valvular disease

Abstract Background In patients with mixed aortic valve disease (MAVD), defined as the combination of aortic stenosis (AS) and aortic regurgitation (AR), the assessment of severity of each valve lesion is often challenging, as most echocardiographic parameters tend to be inaccurate in this setting. There is limited and contradictory data regarding which parameter should be used to risk stratify these patients. Aortic valve area (AVA) is considered a reliable marker of AS severity in the presence of AR, while vena contracta (VC) width provides an accurate estimation of AR severity in the presence of AS. Purpose To evaluate the prognostic value of combining VC and AVA index (AVAi) in asymptomatic patients with significant MAVD (≥ moderate AS and ≥ moderate AR) and left ventricular ejection fraction (LVEF) 50%. Methods A total of 601 patients (mean age 64±17 years, 54% male) with significant MAVD were included. Measurements of VC and AVAi were combined in the VC/AVAi ratio. The study endpoint was all-cause mortality. Aortic valve replacement (AVR) was accounted for during follow-up as a time-dependent covariate. ROC analysis was performed to identify the optimal threshold for predicting mortality, further assessed with restricted cubic spline analysis (Figure 1). Results Over a median follow-up of 7.3 years (IQR, 2.1-12.5), 183 patients (30%) died. The VC/AVAi threshold associated with mortality was ≥10 (Figure 1). Patients with VC/AVAi ≥10 were older (66±16 vs. 60±18 years, p0.001), more frequently male (56% vs. 47%, p=0.013), had lower prevalence of bicuspid aortic valve (BAV, 18% vs. 36%, p0.001), higher relative wall thickness (RWT, 0.5±0.1 vs. 0.4±0.1, p=0.006), and more right ventricular dysfunction (TAPSE 17 mm, 11% vs. 4%, p=0.018). However, comorbidity profiles were similar between patients with VC/AVAi ≥10 and 10. Unadjusted 10-year survival was significantly lower in patients with VC/AVAi ≥10 (62% vs. 84%, p0.001) (Figure 2A). The VC/AVAi ratio remained significantly associated with mortality after adjusting for age, diabetes, coronary artery disease, arterial hypertension, eGFR, BAV, left atrial volume index (LAVI) 34 ml/m2, RWT 0.42, significant tricuspid regurgitation, TAPSE 17 mm, and AVR as a time-dependent covariate (VC/AVA ≥10: HR 2.906, 95%CI 1.764-4.787; p0.001); and with worse adjusted 10-year survival compared to patients with VC/AVAi 10 (70% vs. 86%, p0.001; Figure 2B). Adding VC/AVAi ≥10 to a baseline model that included clinical and echocardiographic variables associated with the endpoint in Cox regression analysis, significantly improved mortality prediction (Chi-square difference=17, p0.001; C-statistic=0.711, NRI=0.408). Moreover, the inclusion of VC/AVAi ≥10 demonstrated the greatest increase in model predictivity compared to the addition of AVAi or VC alone (p0.001). Conclusion The VC/AVAi ratio serves as a more accurate tool for risk stratification in MAVD when compared to VC width or AVAi alone.