Exome risk score for predicting susceptibility to and severity of isolated thoracic aortic aneurysm

Abstract

Abstract Isolated thoracic aortic aneurysms (TAAs) are asymptomatic before dissection or rupture and heterogeneous in clinical phenotype. It is urgent need but difficult to identify individuals at high risk to enable enhanced screening or preventive therapies. Because TAAs have a genetic component, one possible approach is to stratify individuals based on inherited DNA variations. Here, we constructed an integrated exome risk score (ERS) based on both common and rare variants found in whole-exome sequencing through a machine-learning framework in discovery population consisting of 551 cases and 1071 controls. We evaluated the performance of the ERS in an independent population including 151 cases and 779 controls with a raw odds ratio (OR) per 1 standard deviation (SD) = 1.95 and area under the receiver operating characteristic curve (AUC) = 0.680. When adjusted by gender and the first four principal components, OR per SD = 1.68 and AUC reached 0.783. Individuals in the top 20% of ERS distribution had an OR of 3.20 compared with others. Finally, we found that individuals with top 20% ERS developed TAA at a younger age (P = 0.002) and with a larger diameter (P = 0.016) compared with lower ERS, and were more likely to suffer from aortic root aneurysms (P = 0.009). Our analysis provides a global view of the genetic components of isolated TAA. The exome score developed and evaluated here is the first polygenic risk score for TAA and is a promising predictor of disease risk and severity, which will facilitate the implementation of the risk-reduction strategies.