Site-specific variation in radiologic arterial age and its association with postoperative cardiovascular risk

Vascular aging is a key driver of age-related diseases, especially cardiovascular disease, significantly impacting morbidity and mortality. However, aging rates and disease susceptibility vary widely across organs and vascular regions, reflecting substantial heterogeneity. A universally accepted definition of vascular aging remains elusive.

This study aims to identify timely, non-invasive organ-specific imaging markers to facilitate the accurate estimation of vascular age (represented by arterial age) and enhance the prediction of future cardiovascular disease risk.

A total of 669 patients (aged 46 to 90 years) from a biobank were included in the present study. All patients had undergone a computed tomography (CT) angiography prior to carotid endarterectomy and the CT images were used to segment the aorta, the left and right common carotid arteries, the left and right subclavian arteries, as well as the brachiocephalic trunk. Radiomic features describing the segmented arteries were generated. Arterial ages for the six arteries were estimated using radiomic features and elastic net regression in a nested cross-validation setting. Associations between the constructed arterial ages, clinical factors, circulating proteins quantified using a Proseek Multiplex cardiovascular disease panel, cerebrovascular symptoms at baseline, and postoperative cardiovascular events over a 10-year follow-up period, were examined.

Among the arterial ages derived from six arterial territories, the aorta showed the strongest correlation with the chronological age. Hypertension and creatinine levels correlated with arterial ages across all six vascular territories, whereas CRP was only associated with an increased arterial age derived from the right subclavian artery. Eighteen circulating proteins, including growth/differentiation factor 15 and osteoprotegerin, were linked to either chronological age or arterial age. The arterial ages from the aorta, the left and right subclavian arteries, and the combined six arteries were associated with postoperative cardiovascular events. Among them, patients with the highest tertile of right subclavian arterial age had a 4.6-fold higher risk (95% CI 1.8–12.0) of developing postoperative cardiovascular events. No significant association was observed between chronological age and postoperative cardiovascular events. Incorporating arterial age from the right subclavian artery significantly improved postoperative cardiovascular risk prediction, increasing the Concordance-index by 0.05 compared to using chronological age in the model.

Arterial age from six major arteries were associated with clinical risk factors and improved the prediction of postoperative events. This suggests that incorporating arterial age into risk prediction enhances the ability to identify individuals at high cardiovascular risk, particularly among those undergoing carotid endarterectomy surgery.