Page 20 - Journal of Structural Heart Disease Volume 5, Issue 3

59
Original Scientific Article
Table 4. Subgroup analysis of the correlation among AVAs mea- sured by CCT and TTE
ficulty in drawing inner margin of aortic valve cusps because of blooming artifact from severe calcifica- tion, which can lead to inaccurate measurement that may contribute to under- or over- estimation of AVA when compared to absent or low Agatston aortic valve calcium score. We also found that the qualita- tive degree of aortic valve and annular calcification was not associated with difference between AVACCT and AVATTE and AVACCT was not correlated with AVATTE in group with absence or mild grade of aortic valve calcification. These results may indicate that quanti- tative assessment of aortic valve calcification on CCT is a factor that affects the discrepancy between AVATTE and AVACCT.
Of the two TTE parameters, only transvalvular mean pressure gradient was inversely associated with dif- ference between AVACCT and AVATTE in a simple linear regression analysis. There was no correlation between AVACCT and AVATTE in patients with LVEF < 50% or with transvalvular pressure gradient ≤ 40 mmHg. These results do not explain why TTE parameters might have contributed to correlation between AVACCT and AVATTE. However, the low flow state in which AS se- verity is overestimated due to incomplete opening of the calcified aortic valve might result in no correlation between AVACCT and AVATTE in severe AS [20].
AF is common in patients with AS [24]. In this study, 43% of patients had AF. AF may hamper pre- cise measurement of aortic valve hemodynamics on TTE and deteriorate CCT image quality for the assess- ment of AVA due to mis-registration artifacts related to inconsistent RR intervals [26]. With the use of du- al-source CT, diagnostic image quality was obtained for all patients, even for patients with AF. AF was not associated with difference between AVACCT and AVATTE in a simple linear regression analysis. In the subgroup analyses, it was shown that the correlation between AVAs measured by CCT and TTE varied significantly according to presence/absence of AF. As expected, there was no correlation between AVACCT and AVATTE in patients with AF.
Age showed a weak inverse association with dif- ference between AVACCT and AVATTE. This is consistent with previous study [15]. The evaluation of AS in the elderly may be difficult not only because of underly- ing diseases and clinical conditions, but also because of insufficient compliance with imaging testing.
Pearson's r
p-value
Overall (127) LVEF
<50% (38)
≥50% (89)
AVC Agatston score
≤ 1651 (103)
>1651 (24) LVOT Eccentricity
< 0.78 (58)
≥ 0.78 (69) Atrial fibrillation
Yes (54)
No (73) AVC Grade
Significant (98)
Insignificant (29)
Annular Calcification Grade
Significant (41)
Insignificant (86) Transvalvular Mean PG > 40 mmHg (82)
≤ 40 mmHg (45)
0.04 0.82 0.39 <0.001
0.31 0.00 0.17 0.44
0.40 0.002 0.14 0.26
0.04 0.76 0.37 0.001
0.39 <0.001 0.35 0.17
0.33 0.00 0.23 0.02
0.34 0.002 0.04 0.78
AVA = aortic valve area; AVC = aortic valve calcification; CCT = cardiac com- puted tomography; LVEF = left ventricular ejection fraction; LVOT = left ven- tricular outflow track; PG = pressure gradient; TTE = transthoracic echocar- diography
CCT allows for accurate detection, localization, and quantification of calcification of the aortic valve and annulus [23]. We adopted a threshold of 1,651 Ag- atston score which correctly differentiated patients with severe AS from non-severe AS in the setting of low-flow grade AS [24]. Our result was consistent with a previous study showing that numeric difference be- tween AVACCT and AVATTE was reduced with increasing Agatston score [15]. AVACCT was not correlated with AVATTE in group with Agatston aortic valve score > 1,651. High Agatston aortic valve score results in dif-
Ko S. M. et al.
Aortic Valve Area Measured with CT and TTE