LAUSR

Risk of aortic dissection in ascending thoracic aortic aneurysms is not sufficiently captured by size-based metrics. From a biomechanical perspective, dissection may be initiated when wall stress exceeds wall strength. Our objective was to assess the association between aneurysm peak wall stresses and 3-year all-cause mortality. Finite element analysis was performed in 273 veterans with chest computed tomography for surveillance of ascending thoracic aortic aneurysms. Three-dimensional geometries were reconstructed and models developed accounting for pre-stress geometries. A fiber-embedded hyperelastic material model was applied to obtain circumferential and longitudinal wall stresses under systolic pressure. Patients were followed up to three years following the scan to assess aneurysm repair and all-cause mortality. Fine-Gray sub-distribution hazards were estimated for all-cause mortality based on age, aortic diameter, and peak wall stresses, treating aneurysm repair as a competing risk. When accounting for age, sub-distribution hazard of mortality was not significantly increased by peak circumferential stresses (p=0.30) but was significantly increased by peak longitudinal stresses (p=0.008). Aortic diameter did not significantly increase sub-distribution hazard of mortality in either model (circumferential model: p=0.38; longitudinal model: p=0.30). The effect of peak longitudinal stresses on sub-distribution hazard of mortality was maximized at a binary threshold of 355kPa, which captured 34/212(16%) patients with diameter <5cm, 11/36(31%) at 5.0-5.4cm, and 11/25(44%) at ≥5.5cm. Aneurysm peak longitudinal stresses stratified by age and diameter were associated with increased hazard of 3-year all-cause mortality in a veteran cohort. Risk prediction may be enhanced by considering peak longitudinal stresses.