Validation of aortic valve calcium quantification thresholds measured by computed tomography in Asian patients with calcific aortic stenosis

Abstract

Aims Sex-specific thresholds of aortic valve calcification (AVC) have been proposed and validated in Caucasians. Thus, we aimed to validate their accuracy in Asians. Methods and results Patients with calcific aortic stenosis (AS) from seven international centres were included. Exclusion criteria were ≥moderate aortic/mitral regurgitation and bicuspid valve. Optimal AVC and AVC-density sex-specific thresholds for severe AS were obtained in concordant grading and normal flow patients (CG/NF). We included 1263 patients [728 (57%) Asians, 573 (45%) women, 837 (66%) with CG/NF]. Mean gradient was 48 (26–64) mmHg and peak aortic velocity 4.5 (3.4–5.1) m/s. Optimal AVC thresholds were: 2145 Agatston Units (AU) in men and 1301 AU in women for Asians; and 1885 AU in men and 1129 AU in women for Caucasians. Overall, accuracy (% correctly classified) was high and comparable either using optimal or guidelines’ thresholds (2000 AU in men, 1200 AU in women). However, accuracy was lower in Asian women vs. Caucasian women (76–78% vs. 94–95%; P < 0.001). Accuracy of AVC-density (476 AU/cm2 in men and 292 AU/cm2 in women) was comparable to absolute AVC in Caucasians (91% vs. 91%, respectively, P = 0.74), but higher than absolute AVC in Asians (87% vs. 81%, P < 0.001). There was no interaction between AVC/AVC-density and ethnicity (all P > 0.41) with regards to AS haemodynamic severity. Conclusion AVC thresholds defining severe AS are comparable in Asian and Caucasian populations, and similar to those proposed in the guidelines. However, accuracy of AVC to identify severe AS in Asians (especially women) is sub-optimal. Therefore, the use of AVC-density is preferable in Asians.

Revisiting the Aortic Valve Calcium Score in Evaluating the Severity of Aortic Stenosis in Japanese Patients ― A Single-Center Study ―

“…Recently, Guzzetti et al compared AVC thresholds in Asian and Caucasian populations and concluded that AVC density performed better than AVC in terms of diagnostic accuracy. 23 This is in contrast to the results of the present study; however, Japanese patients were not included in that study. We believe that the adoption of AVC throughout the international guidelines with the same thresholds would be more comprehensive.…”

Effect of race on pressure recovery adjustment for prevention of aortic stenosis grading discordance

“…Another potentially useful imaging technique for decreasing the frequency of discordant grading is to use AVC on non-contrast multidetector CT, which has emerged as a highly valuable tool for assessing anatomical AS severity and has been incorporated in the latest guidelines 13–15. In a recent paper, we reported that among 312 patients with discordant echocardiographic grading, 158 (50.6%) were deemed to have true severe AS when using sex-specific AVC density thresholds without a significant racial difference in the prevalence of true severe AS 22. In the current study, sex-specific AVC thresholds significantly decreased the frequency of discordant grading in both races and the simultaneous use of both sex-specific AVC thresholds and PR adjustment further decreased the frequency of discordant grading.…”

Aortic Valve Calcification Density Measured by MDCT in the Assessment of Aortic Stenosis Severity

“…14,15 Therefore, as interindividual valve size variability is not considered, the current method may lead to an overestimation or underestimation of AS true severity, especially in patients with eccentric, large annuli (ie, patients with bicuspid aortic valve) or very small annuli (smaller patients). [16][17][18] Thus, substantial improvement could potentially be observed in both the diagnosis and management of patients with AS with the implementation of AVCd values in clinical practice as it improves predicting values. 3 The rounded thresholds could be 300 AU/cm 2 in women and 500 AU/cm 2 in men since they provide a similar classification of AS severity for patients in our cohort (data not shown).…”