Age-related changes of thoracic aorta geometry used to predict the risk for acute type B dissection

Abstract

Aims: Risk models that use a single aortic diameter threshold have failed to successfully predict acute type B aortic dissection (TBAD). We sought to identify meaningful age-indexed anatomical variables to predict TBAD risk. Methods and results: A geometric deformable model, consisting of virtual elastic balloons that inflate inside a vessel lumen, was developed to quantify thoracic aorta geometry. In the presence of TBAD, true and total artery lumen morphology were assessed. A stepwise logistic model was built to predict TBAD risk. Initial covariates included age, gender, body mass index and all anatomic variables not directly related to the dissected segment. Patients with acute TBAD (n = 34, 62 ± 12 years old, 57% male gender) were compared with subjects with symptoms of dissection, but with a subsequent negative diagnosis (n = 51, 62 ± 12 years old, 76% male gender). Patient risk factors did not differ between groups. Most aortic anatomical variables were age-dependent. Aortic size was larger in every segment of the dissected with respect to non-dissected aortas (p < 0.001). Variables entering the TBAD risk prediction model were aortic arch diameter, thoracic aorta length and age (predictability = 0.9764, r = 0.85), confirmed by a bootstrap internal validation. In dissected aortas, the true lumen volume was correlated to age (r = 0.72). Conclusions: TBAD probability increases with a larger aortic arch diameter and a longer thoracic aorta, whereas threshold values increase with age. The aortic morphology was age-dependent. After dissection, true lumen volume correlated to age. The use of threshold values indexed to age should be encouraged to better prevent and eventually treat TBAD.