Abstract P112: Prediction of the Development of Aortopathy in Patients With Bicuspid Aortic Valves

Background: Little is known about the association between bicuspid aortic valves (BAV) and ascending aorta dilatation in the population. We aimed to firstly determine the predictors of aortopathy in BAV patients and secondly to develop a clinical classifier to predict aortopathy in a large group of individuals with bicuspid aortic valve. Method: This study includes 1041 patients (including 545 BAV patients and 506 tricuspid aortic valves (TAV) patients) aged 18 or above with aortic valve disease and/or ascending aorta dilatation but devoid of coronary artery disease and primarily not planned for another concomitant valve surgery. Aortic complications of patients were assessed at baseline. Using automated machine-learning, we applied 10-fold cross-validation logistic regression incorporating multidimensional information (i.e., valvular dysfunction, valves morphology, blood sampling, genetic data, clinical data, family history of cardiovascular diseases, prevalent diseases data, demographic characteristics, lifestyle habits data and medication). Results: Among the 545 BAV subjects (age 64.89 +/- 12 years, 68% men). The prevalence of BAV associated with aneurysm (dilatation) (BAV-D) and without aneurysm (BAV-ND) was 54.9% and 45.1% (p<0.001), respectively. Comparing BAV-D and BAV-ND, significant differences in valvular dysfunction pattern were noted, with aortic stenosis predominating in BAV patients without aneurysm (79.3% vs. 68.7% in BAV with aneurysm; p=0.007), and aortic regurgitation in BAV patients with aneurysm (33.7% vs. 25.1% in BAV patients without aneurysm; p =0.03). No differences in age, prevalence of male sex and BMI were observed between BAV-D and BAV-ND Our descriptive analysis showed several patterns of significantly differently associated traits, between undilated and dilated BAV patients (i.e, blood pressure, hypertension, body surface area, height, high-sensitivity CRP, sibling history of myocardial infarction and mother history of myocardial infarction before 65 years, low density lipoprotein (LDL), prevalent MI, prevalent diabetes and prevalent stroke, BAV phenotype). Our predictive classifier included these traits and showed sensitivity of 80% specificity of 84%, negative predictive value of 76%, and positive predictive value of 82% to predict BAV individuals who are of a high risk of developing aneurysm. All these analyses were also performed in TAV patients and showed different patterns of aneurysm manifestation compared to BAV patients Conclusion: Our findings raise the issue of how to implement prevention of aortopathy in BAV patients in a clinical setting and suggest/demonstrated that cardiovascular risk profiles appear to be more predictive than valve morphology, genetic data and circulating plasma proteins.