Ascending aorta dilatation is common in bicuspid aortic valve (BAV). The aim of this study was to investigate agreement of transthoracic echocardiographic (TTE) measurement of the sinuses of Valsalva and the tubular mid–ascending aorta (Asc-Ao) compared with electrocardiographically gated computed tomographic angiographic (CTA) assessment in patients with BAV.
Fifty-three patients with BAV (mean age, 54 ± 14 years; 74% men) who underwent both TTE and CTA imaging for ascending aortic assessment were retrospectively identified. All studies were measured de novo by experts. TTE measurements were obtained at the sinuses and the Asc-Ao, at both systole and end-diastole, using both leading edge–to–leading edge (L-L) and inner edge–to–inner edge (I-I) methods in the parasternal long-axis (LAX) view. The sinuses were also measured in the parasternal short-axis (SAX) view using the same methods plus mid-diastole. CTA measurements were obtained in diastole using outer wall–to–outer wall (O-O) and inner wall–to–inner wall (I-I) methods. Correlation and agreement between the two imaging modalities were assessed using Lin correlation and Bland-Altman analysis, respectively.
Compared with CTA O-O maximum sinuses diameter, the best correlation and agreement were obtained using the TTE SAX mid-diastolic L-L method (ρ = 0.89, 2.6 ± 2.3 mm, respectively). Compared with CTA O-O maximum Asc-Ao diameter, the TTE LAX systolic L-L method (ρ = 0.93, 1.3 ± 2.5 mm) was best. Compared with CTA I-I maximum sinuses diameter, the TTE SAX mid-diastole L-L method (ρ = 0.95, 0.6 ± 2.2 mm) was unbiased. Compared with CTA I-I maximum Asc-Ao diameter, the TTE LAX end-diastolic L-L method (ρ = 0.95, 0.6 ± 2.4 mm) was unbiased.
In patients with BAV aortopathy, unbiased agreement between CTA and TTE imaging can be obtained between the CTA I-I method and TTE SAX mid-diastolic L-L method for the sinuses and the TTE LAX end-diastolic L-L method for the Asc-Ao. When using the CTA O-O method, the best agreement is obtained with the TTE SAX mid-diastolic L-L method for the sinuses (bias ∼2 mm) and the TTE LAX systolic L-L method (bias ∼1 mm) for the Asc-Ao.
In patients with bicuspid aortic valve aortopathy, there is significant systematic underestimation of sinuses of Valsalva by the conventional TTE end-diastolic L-L method, as compared to ECG-gated CTA.
In patients with bicuspid aortic valve aortopathy, unbiased agreement between CTA and TTE imaging can be obtained between the CTA I-I method and the TTE short-axis mid-diastolic L-L method for the sinuses and the conventional TTE method for the Asc-Ao.
In patients with bicuspid aortic valve aortopathy, when using the CTA O-O method, the best agreement is obtained with the TTE short-axis mid-diastolic L-L method for the sinuses (bias ∼2 mm) and the conventional TTE method for the Asc-Ao (bias ∼1 mm).
The ascending thoracic aorta includes the sinuses of Valsalva and the tubular mid–ascending aorta (Asc-Ao). Dilatation of the ascending thoracic aorta, a manifestation of aortopathy, is critical to recognize and monitor because it predicts aortic dissection. Patients with bicuspid aortic valve (BAV) are at increased risk for aortopathy, with 25% risk for developing ascending aortic aneurysms 25 years after BAV diagnosis and risk for aortic dissection between 0.03% to 0.1% cases per patient per year. This risk may increase to 0.5% per patient per year in those with diameters ≥ 45 mm. Although the sinuses may dilate in patients with BAV, the Asc-Ao is the most common portion affected (60%–70% of cases). Important clinical decisions are based primarily on ascending thoracic aortic diameters in patients with BAV. As such, accurate and reproducible measurements of the ascending aorta are critical.
Currently, the American Society of Echocardiography (ASE) recommends transthoracic echocardiographic (TTE) measurement of the ascending thoracic aorta using the leading edge of the anterior aortic wall to the leading edge of the posterior aortic wall at end-diastole, whereas American College of Cardiology (ACC)/American Heart Association (AHA) guidelines recommend using the inner edge–to–inner edge (I-I) method. Technical issues may affect the accuracy of TTE measurements; the ultrasound beam may be aligned obliquely instead of perpendicular to the long axis (LAX) of the aorta. In addition, the BAV sinuses may dilate asymmetrically, leading to diameter underestimation by single TTE two-dimensional view (parasternal LAX). Also, measurement protocols by edge selection (I-I vs leading edge–to–leading edge [L-L]) and by trigger time (diastole vs systole) may cause systematic measurement variation.
With these potential TTE limitations, advanced imaging modalities such as electrocardiographically gated computed tomographic angiographic (CTA) imaging are recommended to confirm aortic dilation. For CTA imaging, current ACC/AHA guidelines recommend external diameter outer wall–to–outer wall (O-O) measurement. More recently, ASE guidelines have suggested inner wall–to–inner wall (I-I) measurement for the ascending aorta on CTA imaging and O-O measurement for the descending and abdominal aorta.
TTE imaging has been noted to systematically underestimate the sinuses and Asc-Ao dimensions compared with CTA imaging, but other studies have not found significant underestimation. Recently, Rodriguez-Palomares et al. reported the best measurement agreement between the conventional TTE end-diastolic L-L method and the CTA diastolic I-I method for both the sinuses and the Asc-Ao. However, that study had few patients with BAV, and patients with sinus asymmetry were excluded. The aim of the present study was twofold: (1) assess TTE and CTA differences in sinuses and Asc-Ao diameter measurements by recommended guidelines in patients with BAV, and (2) explore alternative TTE and CTA measurement methods to reconcile these differences.
Consecutive adult patients (age ≥ 18 years) with BAV ( N = 53) who underwent both TTE and CTA imaging for aortic assessment within 6 months were retrospectively identified between 2006 and 2012. All TTE and CTA studies were measured de novo by experts. The study protocol was approved by the institutional review board.
TTE studies were performed on commercially available ultrasound systems (Vivid 7 [GE Healthcare, Milwaukee, WI] and iE33 [Philips Medical Systems, Andover, MA]). Studies were remeasured offline at both end-diastole and where best seen during systole using the two methods of edge selection in the parasternal LAX and short-axis (SAX) views ( Figure 1 ): (1) sinuses (LAX and SAX) and (2) Asc-Ao (LAX only). For the sinuses SAX measurement, the outer leading edge chosen was the right coronary sinus, which is perpendicular to the ultrasound beam (and thus has the best lumen-wall interface definition), measuring sinus to sinus (right to non and right to left), followed by a third non-to-left sinus measurement ( Figure 1 ). An additional “timing of measurement” was used for SAX sinuses measurement only: mid-diastole ( Figure 1 ). The Asc-Ao was defined as the largest clearly visualized dimension at approximately ≥2 cm above the sinotubular junction. To evaluate interobserver agreement, measurements were repeated by a second expert reader on a randomly selected subset of 19 patients. The term conventional method refers to TTE LAX L-L aortic measurement in end-diastole.
CTA imaging of the chest was performed using multislice dual-energy computed tomographic scanners (Siemens, Forchheim, Germany). Axial images 1 to 2 mm thick, reconstructed in a diastolic phase, were exported into dedicated reformatting software (Aquarius Intuition; TeraRecon, Foster City, CA) and used for analysis. The locations of maximum caliber of the sinuses and Asc-Ao were determined using multiplanar reformatting. A SAX image of the Asc-Ao perpendicular to the direction of blood flow was made using double-oblique reformatting at the site of maximum Asc-Ao caliber. Measurements of the maximum Asc-Ao diameter were made on this reformatted SAX image. A second measurement of the Asc-Ao caliber 90° perpendicular to the measurement of maximum diameter was also made on the SAX image ( Figure 2 ). Thus, Asc-Ao measurements were (1) maximum and (2) perpendicular to maximum.
A SAX image of the sinuses perpendicular to the direction of blood flow was made using double-oblique reformatting at the site of maximum sinuses caliber. The maximum sinuses diameter was measured on this SAX image, from sinus to sinus. The sinuses were also measured in a three-chamber view of the left ventricle (analogous to the TTE parasternal LAX view) created using multiplanar reformatting ( Figure 2 ). Thus, sinuses measurements were (1) maximum and (2) three-chamber view.
Measurements were made using the O-O method as per institutional protocol but also the I-I method for research purposes. To evaluate interobserver agreement, measurements were repeated by a second expert reader on a randomly selected subset of 19 patients.
Continuous data are expressed as mean ± SD with ranges provided when appropriate. Comparisons between measurements were assessed using Lin concordance correlation coefficients and Bland-Altman plots. Assessment of interobserver variability was performed offline on the same images. Statistical analyses were performed using SAS version 9.3 (SAS Institute, Cary, NC) and R version 3.0.2 (R Foundation for Statistical Computing, Vienna, Austria).
Baseline characteristics are shown in Table 1 . Of 53 patients, 39 were men (74%). The mean age was 54 ± 14 years (range, 25–82 years). Forty patients (75%) had fusion of the right and left coronary cusps, 10 (19%) fusion of the right and noncoronary cusps, and one (2%) had fusion of the left and noncoronary cusps. Two patients (4%) were subsequently determined to have unicuspid aortic valves but were included because they were initially identified as having BAV, and unicuspid patients share significant natural history similarities with those with BAV, including aortopathy. Thirty-four patients (64%) had aortic stenosis, while 38 patients (72%) had aortic regurgitation. Mean time from TTE imaging to CTA imaging was 12 ± 29 days.
|Age (y)||54 ± 14|
|Heart rate (beats/min)||69 ± 13|
|Systolic blood pressure (mm Hg)||122 ± 18|
|Diastolic blood pressure (mm Hg)||74 ± 11|
|Weight (kg)||92 ± 20|
|Height (cm)||175 ± 10|
|Body mass index (kg/m 2 )||29.8 ± 5.8|
|Body surface area (m 2 )||2.1 ± 0.2|
|Ejection fraction (%)||63 ± 7|
|Left ventricular mass (g)||238 ± 82|
|Bicuspid ascending aortic valve morphology|
|Type 1 (fusion of right and left coronary cusps)||40 (75)|
|Type 2 (fusion of noncoronary and right coronary cusps)||10 (19)|
|Type 3 (fusion of noncoronary and left coronary cusps)||1 (2)|
|Aortic valve calcification||31 (58)|
|Aortic valve stenosis|
|Aortic valve regurgitation|
Aortic Measurements by Imaging Modality and Measuring Method
Assessment of the Asc-Ao and sinuses LAX by TTE imaging in systole allowed the best qualitative visualization of their dimensions (compared with end-diastole); nonetheless, all 53 patients had measurable images by either method. Assessment of the sinuses SAX by TTE imaging at mid-diastole allowed the best qualitative visualization of their SAX dimension (46 patients [87%] had measureable images), compared with systole (42 [79%]) and end-diastole (36 [68%]).
All systolic L-L Asc-Ao and sinuses LAX diameters were larger than those obtained at end-diastole by either the I-I or the L-L method ( P < .0001), with the exception of the mid-diastole L-L sinuses SAX, which was larger than the all sinuses LAX measurements ( P < .001) ( Table 2 ). The maximum aortic diameters measured by TTE imaging were systolic L-L Asc-Ao (43.8 ± 7.5 mm) and mid-diastolic L-L sinuses SAX (42.5 ± 7 mm). By the conventional measurement method (LAX end-diastolic L-L), the sinuses were dilated (≥45 mm) in 8 patients (15%), which increased to 11 patients (21%) by LAX systolic L-L measurement and to 16 (35%) by SAX mid-diastolic L-L measurement ( Table 3 ). Similarly but less marked, by the conventional method, the Asc-Ao was dilated (≥45 mm) in 26 patients (49%), which increased to 27 patients (51%) with LAX systolic L-L.
|Mean ± SD||Mean ± SD of paired difference||P|
|Mid-diastole||End-diastole||Systole||Mid-diastole − end-diastole||Mid-diastole − systole||End-diastole − systole|
|Sinuses LAX L-L||39.4 ± 6.8||40.4 ± 6.9||−1 ± 1.4||<.0001|
|Sinuses LAX I-I||37.7 ± 6.8||38.7 ± 6.9||−1 ± 1.3||<.0001|
|Sinuses SAX L-L||42.5 ± 7.0||40.7 ± 6.7||41.8 ± 7.4||1.2 ± 2||−0.1 ± 1||−1.2 ± 1.6||<.0001|
|Sinuses SAX I-I||39.1 ± 6.7||39.0 ± 7.2||−0.4 ± 0.9||.00187|
|Asc-Ao L-L||42.6 ± 7.4||43.8 ± 7.5||−1.2 ± 1.6||<.0001|
|Asc-Ao I-I||41.1 ± 7.4||42.3 ± 7.3||−1.2 ± 1.7||<.0001|
|Mean ± SD||Mean ± SD of paired difference||P|
|Maximum||Three-chamber or perpendicular||Maximum − three-chamber/perpendicular|
|Sinuses O-O||44.7 ± 7.0||40.5 ± 6.9||4.3 ± 2.4||<.0001|
|Sinuses I-I||42.6 ± 7.0||38.2 ± 6.8||4.4 ± 2.3||<.0001|
|Asc-Ao O-O||45.2 ± 7.6||44.0 ± 7.7||1.2 ± 0.9||<.0001|
|Asc-Ao I-I||43.3 ± 7.5||41.8 ± 7.5||1.4 ± 0.9||<.0001|
|Sinuses diameter < 40 mm|
|Systole LAX L-L||53||24 (45)|
|End-diastole LAX L-L||53||28 (53)|
|Systole LAX I-I||53||30 (57)|
|End-diastole LAX I-I||53||38 (72)|
|Systole SAX L-L||42||16 (38)|
|End-diastole SAX L-L||36||14 (39)|
|Mid-diastole SAX L-L||46||14 (30)|
|Systole SAX I-I||42||24 (57)|
|Mid-diastole SAX I-I||46||27 (59)|
|CTA three-chamber O-O||52||26 (50)|
|CTA three-chamber I-I||52||31 (60)|
|CTA maximum O-O||52||11 (21)|
|CTA maximum I-I||52||16 (31)|
|Sinuses dilation ≥ 45 mm|
|Systole LAX L-L||53||11 (21)|
|End-diastole LAX L-L||53||8 (15)|
|Systole LAX I-I||53||8 (15)|
|End-diastole LAX I-I||53||7 (13)|
|Systole SAX L-L||42||14 (33)|
|End-diastole SAX L-L||36||11 (31)|
|Mid-diastole SAX L-L||46||16 (35)|
|Systole SAX I-I||42||7 (17)|
|Mid-diastole SAX I-I||46||7 (15)|
|CTA three-chamber O-O||52||15 (29)|
|CTA three-chamber I-I||52||8 (15)|
|CTA maximum O-O||52||28 (54)|
|CTA maximum I-I||52||18 (35)|
|Mid–ascending aorta diameter < 40 mm|
|Systole LAX L-L||53||19 (36)|
|End-diastole LAX L-L||53||22 (42)|
|Systole LAX I-I||53||22 (42)|
|End-diastole LAX I-I||53||23 (43)|
|CTA perpendicular O-O||52||15 (29)|
|CTA perpendicular I-I||52||23 (44)|
|CTA maximum O-O||52||11 (21)|
|CTA maximum I-I||52||19 (37)|
|Mid–ascending aorta dilation ≥ 45 mm|
|Systole LAX L-L||53||27 (51)|
|End-diastole LAX L-L||53||26 (49)|
|Systole LAX I-I||53||24 (45)|
|End-diastole LAX I-I||53||21 (40)|
|CTA perpendicular O-O||52||25 (48)|
|CTA perpendicular I-I||52||23 (44)|
|CTA maximum O-O||52||28 (54)|
|CTA maximum I-I||52||26 (50)|