Abstract
Background: In patients with aortic stenosis (AS), measurement of aortic valve calcification (AVC) using computed tomography (CT) is recommended in cases where echocardiographic measurements are inconclusive. However, sex-specific AVC thresholds proposed in the guidelines for predicting severe AS (women: 1,200 arbitrary units [AU]; men: 2,000 AU) are based on studies from Western countries. Methods and Results: We retrospectively included 512 Japanese patients with at least moderate AS who underwent transthoracic echocardiography and CT. AVC was quantified using the Agatston method. AVC was positively correlated with peak aortic jet velocity and mean transvalvular gradient (mPG), and negatively correlated with aortic valve area (AVA) and the AVA index (AVAi). In 257 patients with concordant AS grading (152 severe AS [AVAi ≤0.6 cm(2)/m(2), mPG ≥40 mmHg], 105 moderate AS [AVAi >0.6 cm(2)/m(2), mPG <40 mmHg]), receiver operating characteristic curve analysis of AVC predicting severe AS yielded an area under the curve of 0.91 (95% confidence interval [CI] 0.87-0.95; P<0.001) in women and 0.86 (95% CI 0.75-0.98; P<0.001) in men. The optimal thresholds (women: 1,379 AU; men: 1,802 AU) were close to those proposed in the guidelines. The diagnostic accuracy of the thresholds in the guidelines was similar to that of the optimal thresholds. Conclusions: The sex-specific AVC thresholds proposed in international guidelines can be applied to Japanese AS patients, yielding similar diagnostic accuracy as the optimal cut-off derived from the study patients.