Quantitative feasibility of aortic-valve agatston score derived from 5 ​mm-thick non-electrocardiography-gated noncontrast body computed tomography for evaluating severe aortic stenosis.

Abstract

Background: The aortic-valve Agatston score (AVAS) is valuable for evaluating severe aortic stenosis (AS). While visual assessment of AS using chest computed tomography (CT) during lung cancer screening facilitates qualitative evaluation, it remains unclear whether AVAS derived from body CT that are neither electrocardiography (ECG)-triggered nor ECG-gated can quantitatively evaluate severe AS. This study aims to investigate the quantitative feasibility of AVAS derived from the 5 ​mm-thick noncontrast body CT for evaluating severe AS.

Methods: In this retrospective study, data were collected from participants who underwent both cardiac CT scans that were either ECG-gated or ECG-triggered and noncontrast body CT scans that were neither ECG-triggered nor ECG-gated prior to AS treatment. We quantified AVAS from the body CT scan with a slice thickness of 5 ​mm (body CT AVAS) and AVAS from the cardiac CT scan with a slice thickness of 3 ​mm (cardiac CT AVAS). Regression analysis was performed between body CT AVAS and cardiac CT AVAS. Receiver-operating characteristic (ROC) curve analysis of body CT AVAS was conducted to detect cardiac CT AVAS of ≥2000 and ​≥1300.

Results: A total of 265 participants (90 males; median age, 84 years [interquartile range, 80-88 years]) were analyzed. Regression analysis between body CT AVAS and cardiac CT AVAS yielded an R² of 0.92. Body CT AVAS of 2540 and 1440 corresponded to cardiac CT AVAS of 2000 and 1300, respectively. The areas under the ROC curves were 0.99 and 0.98, respectively.

Conclusion: Five mm-thick noncontrast body CT AVAS is a quantitatively feasible tool for evaluating severe AS.