The authors hypothesized that aortic root geometry is different between bicuspid and tricuspid aortic stenosis (AS) that can be assessed using real-time three-dimensional (3D) transesophageal echocardiography. The aims of this study were (1) to validate the accuracy of 3D transesophageal echocardiographic measurements of the aortic root against multidetector computed tomography as a reference, (2) to determine the difference of aortic root geometry between patients with tricuspid and bicuspid AS, and (3) to assess its impact on pressure recovery.
In protocol 1, 3D transesophageal echocardiography and contrast-enhanced multidetector computed tomography were performed in 40 patients. Multiplanar reconstruction was used to measure the aortic annulus, the sinus of Valsalva, and the sinotubular junction area, as well as the distance and volume from the aortic annulus to the sinotubular junction. In protocol 2, the same 3D transesophageal echocardiographic measurements were performed in patients with tricuspid AS (n = 57) and bicuspid AS (n = 26) and in patients without AS (n = 32). The energy loss coefficient was also measured in patients with AS.
In protocol 1, excellent correlations of aortic root geometric parameters were noted between the two modalities. In protocol 2, compared with patients without AS, those with tricuspid AS had smaller both sinotubular junction areas and longitudinal distances, resulting in a 23% reduction of aortic root volume. In contrast, patients with bicuspid AS had larger transverse areas and longitudinal distances, resulting in a 30% increase in aortic root volume. The energy loss coefficient revealed more frequent reclassification from severe AS to moderate AS in patients with tricuspid AS (17%) compared with those with bicuspid AS (10%).
Three-dimensional transesophageal echocardiography successfully revealed different aortic root morphologies between tricuspid and bicuspid AS, which have different impacts on pressure recovery.
With growing aging populations in industrialized nations, symptomatic patients with severe calcific tricuspid aortic stenosis (AS) with high surgical risk are more likely to be referred for transcatheter aortic valve replacement (TAVR). The accurate determination of aortic root geometry before TAVR is related to the rate of successful procedures without complications.
Multidetector computed tomography (MDCT) has been the preferred method for the evaluation of aortic root geometry in patients with tricuspid AS who are candidates for TAVR. However, controversy exists regarding aortic root remodeling in tricuspid AS using MDCT. Because of the concern of radiation exposure, risk for contrast-induced nephropathy, relative contraindication of atrial fibrillation, and contraindication of hypersensitivity to iodine contrast agents, three-dimensional (3D) transesophageal echocardiography (TEE) has emerged and been demonstrated reliably to evaluate aortic root geometry as an alternative or complement to MDCT.
Although the severity of AS is usually determined by aortic valve area (AVA) using the continuity equation, there is a risk for overestimation if correction for pressure recovery is not performed, especially in patients with small aortas.
We hypothesized that aortic root geometry is different between tricuspid AS and bicuspid AS, which could have a significant impact on pressure recovery and appropriate strategies for TAVR procedures. Accordingly, the aims of this study were (1) to validate the accuracy of 3D transesophageal echocardiographic measurements of aortic root geometry against MDCT as a reference, (2) to compare aortic root geometry between patients with tricuspid and bicuspid AS, and (3) to assess how different changes in aortic root geometry can affect pressure recovery phenomena between the two groups.
In protocol 1, we retrospectively enrolled 40 patients who underwent both coronary MDCT and 3D TEE within 1 month. All patients had received coronary MDCT to diagnose coronary artery stenosis. Patients had also undergone 3D TEE for various indications. Patients who were allergic to iodine contrast agents and those at high risk for contrast nephropathy were excluded. In protocol 2, 57 patients with tricuspid AS and 26 with bicuspid AS who underwent clinically indicated 3D TEE were consecutively enrolled from April 2008 to May 2012. Because it was impossible to enroll strictly normal subjects for 3D TEE as a control group, 32 patients undergoing 3D TEE to determine the source of embolisms (n = 28) or to evaluate cardiac masses in the right heart chamber (n = 4) during the same time window who were subsequently diagnosed to have normally functioning aortic valves were selected as a non-AS group for the comparison of aortic root geometry. The study was approved by the ethics committee at the University of Occupational and Environmental Health hospital, and written informed consent was obtained from all patients at the time of 3D TEE.
The multidetector computed tomographic examinations were performed using a 64–detector row scanner (Aquilion 64; Toshiba Medical Systems, Tokyo, Japan). A nonionic iodinated contrast agent (Iopamiro 300; Bayer Healthcare, Osaka, Japan) at 1 mL/kg (4 mL/sec) was injected into the antecubital vein, followed by a 30-mL saline bolus. Image acquisition was triggered by the appearance of contrast in the aortic root. Imaging parameters included a gantry rotation time of 350 msec with 5 mm per rotation, tube voltage of 120 kV, and tube current of 500 mA. Scan data were then reconstructed at a slice thickness of 0.5 mm with 0.5-mm slice resolution using retrospective electrocardiographic gating from end-diastole (0% of the RR interval) to late diastole (90% of the RR interval) at 10% steps. All patients had taken an oral β-blocker (atenolol 25 mg) in the morning of the day of MDCT.
Three-dimensional TEE was performed using a commercially available ultrasound imaging system (iE33; Philips Medical Systems, Andover, MA) with a 3D matrix-array transesophageal transducer (X7-2t; Philips Medical Systems). After induction of topical pharyngeal anesthesia and intravenous sedation, the transesophageal echocardiographic probe was advanced into the esophagus. From the midesophageal position, a long-axis two-dimensional (2D) view (135°) of the aortic valve was obtained. The 3D zoom mode, which displays a smaller magnified pyramidal volume, was subsequently activated to image the aortic root. Using a biplane image, the size of the pyramidal box was adjusted to ensure that the entire aortic root from the aortic annulus to the sinotubular junction (STJ) were included in the scan volume. Gain and compression, as well as time gain compensation, were optimized. Three-dimensional zoom data sets with one-beat acquisition during two consecutive cardiac cycles were stored digitally.
2D Transthoracic Echocardiography
Because the highest jet velocity across the aortic valve in patients with AS is not usually obtained using TEE, we obtained continuous-wave Doppler–derived peak velocity across the aortic valve and pulsed-wave Doppler velocity at the left ventricular outflow tract on 2D transthoracic echocardiography performed as close in time as possible to 3D TEE (median examination time interval, 7 days; range, 0–68 days). To measure peak velocity, multiple locations of continuous-wave flow velocity measurements were performed to acquire the highest velocity across the aortic valve.
MDCT Measurements of the Aortic Root
Measurements of the aortic root parameters were performed in the coronal, single-oblique sagittal, and double-oblique transverse views extracted from 3D multidetector computed tomographic data sets, as previously described. For measurements of AVA, the frame showing maximal opening of the aortic valve was selected (usually the frame at 20% or 30% of the RR interval), and AVA was manually traced in the double-oblique transverse view. For measurements of the aortic annulus, sinus of Valsalva, and STJ, the double-oblique transverse view was realigned to be perpendicular to the long axis of the aortic root at the desired level, and its boundaries were manually traced ( Figures 1 A–1C). For the measurement of aortic root volume, 14 equidistant cutting planes orthogonal to the long-axis view were applied from the aortic annulus to the STJ in two long-axis views of the aortic root. The aortic root volume was calculated using a summation of slice area times slice width. The distance from the annulus to the coronary orifice was measured at the end-diastolic frame (0% or 90% of the RR interval) because of easy visualization of the relationship between the two structures. The left coronary artery ostium was identified in the oblique coronal reformation views oriented orthogonally to the plane of the aortic annulus. The right coronary artery ostium was identified on oblique sagittal reformation views. Measurements were performed from the lower part of each ostium to the aortic annulus, perpendicular to the annulus ( Figures 1 D and 1E). These measurements were performed using commercial software (zioTerm2009; Ziosoft Inc, Tokyo, Japan).
3D TEE Measurements of the Aortic Root
From the zoomed 3D data sets, two orthogonal long-axis views of the aortic root (anterior-posterior and medial-lateral projections) were extracted using the multiplanar reconstruction mode ( Figure 2-1 ). A third plane perpendicular to both of the long-axis planes was manipulated to obtain the transverse 2D cutting plane of the aortic valve. After choosing the midsystolic frame in which maximal aortic valve excursion was observed, fine adjustments of the cutting plane were performed to obtain the smallest aortic valve orifice, from which AVA was manually traced using the magnified view mode ( Figure 2 A, bottom left ) with quantitative software (3DQ, QLAB; Philips Medical Systems). Subsequently, the multiplanar reconstruction planes were aligned at each level of the aortic root (annulus, sinus of Valsalva, and STJ) to obtain the true cross-sectional area. The aortic annular level was defined as the lowest plane of the valve hinge point (inferior virtual basal ring; Figures 2 B–D). For the determination of aortic root volume, 14 equidistant cutting planes were applied from the aortic annulus to the STJ on two long-axis views of the aortic root ( Figure 2 A top , dotted lines ). The aortic root volume was calculated using a summation of slice area times slice width as follows: aortic root volume = Σ A i × slice width, where i = 1 to 13, A is area, and slice width = aortic annulus–to–STJ height/13.
Finally, an end-diastolic frame was selected for the measurement of the distance between the annulus and the coronary ostia. For the measurement of the distance between the aortic annulus and the left coronary ostia, a short-axis view at the level of left coronary ostium was obtained using multiplanar reconstruction mode. The orthogonal plane that dissected at the left coronary ostium was applied, from which the shortest distance between the aortic annulus and the left coronary orifice could be measured ( Figure 2 E). The distance between the aortic annulus and the right coronary orifice was usually obtained in the long-axis view of the aortic root extracted from 3D zoomed data sets ( Figure 2 F).
2D TEE Measurement
AVA by the continuity equation (AVA 2DCE ) was measured using the standard method. Energy loss coefficient (ELCo) was calculated as follows: ELCo = AVA 2DCE × STJ area/(STJ area − AVA 2DCE )
STJ area was determined using 2D transesophageal echocardiographic determination of STJ diameter (the shape of the STJ is assumed to be circular) and direct tracing of STJ area on 3D TEE.
Intraobserver variability was determined by having the observer repeat the measurement of aortic annular area, sinus of Valsalva area, STJ area, distance from the aortic annulus to the STJ, and aortic root volume 1 month later using 3D transesophageal echocardiographic images in 15 randomly selected patients. Interobserver variability was determined by having a second observer perform these measurements in the same 15 patients.
Continuous data are expressed as mean ± SD. Categorical data are expressed as numbers or percentages. All statistical analysis was carried out using commercial software (JMP version 9.0, SAS Institute Inc, Cary, NC; SPSS version 21, SPSS, Inc, Chicago, IL). Categorical variables were compared using Fisher’s exact test or χ 2 test as appropriate. Differences in continuous variables among the three groups were assessed using repeated-measures of analysis of variance with post hoc Tukey-Kramer honestly significantly different correction. Student’s t tests were used to test the differences in continuous variables between the two groups. Linear regression analysis was used to study the relationships between two parameters. Bland-Altman analysis was performed to determine bias and limits of agreement between two measurements. Intraobserver and interobserver variability values were calculated as the absolute differences between the corresponding two measurements as percentages of their mean. Interobserver and intraobserver reproducibility was determined using intraclass correlation coefficients. P values < .05 were considered significant.
Study patients’ demographics are shown in Table 1 . Among 40 patients, 15 had more than moderate AS. Aortic root parameters were successfully measured on MDCT and 3D TEE in all patients. The mean frame rate on 3D TEE for the assessment of the aortic root was 16 ± 7 frames/sec (range, 8–45 frames/sec). Figure 3 shows the linear correlation and Bland-Altman plot between multidetector and 3D transesophageal echocardiographic measurements in each parameter. All root parameters measured by 3D TEE were slightly but significantly smaller compared with those measured on MDCT, but there was good correlation between the two imaging modalities ( r = 0.71–0.96).
|Age (y)||70 ± 13 (26–89)|
|Heart rate (beats/min)||68 ± 12 (49–90)|
|Systolic BP (mm Hg)||150 ± 29 (85–203)|
|Diastolic BP (mm Hg)||79 ± 15 (44–111)|
|BSA (m 2 )||1.53 ± 0.19 (1.24–1.94)|
|Diabetes mellitus||9 (22.5%)|
|Aortic valve stenosis||15 (37.5%)|
|Other valve disease||11 (27.5%)|
|Known CAD||9 (22.5%)|
|Atrial fibrillation||2 (5.0%)|
|LVEDVI (mL/m 2 )||82 ± 29 (42–157)|
|LVESVI (mL/m 2 )||43 ± 22 (17–114)|
|LVEF (%)||49 ± 11 (21–64)|
Detailed baseline clinical characteristics for the three groups are shown in Table 2 . Compared with patients with bicuspid AS, those with tricuspid AS were significantly older and had higher systolic and diastolic blood pressures and higher creatinine levels, resulting in lower estimated glomerular filtration rates. However, indexed AVA using 3D TEE showed no significant differences between the two groups. Standard 2D echocardiographic results are shown in Table 3 . Patients with tricuspid AS had significantly higher E-wave and A-wave velocities and lower E′, resulting in higher E/e′ ratios compared with patients with bicuspid AS. However, peak aortic valve jet velocity and mean pressure gradient were again nearly equivalent. Aortic root characteristics are shown in Table 4 , and geometric renderings of the different sizes of aortic root dimensions among three groups are shown in Figure 4 . All parameters were indexed to body surface area. As stated in previous reports, the aortic annulus showed an ellipsoid shape in all three groups. On the other hand, STJ shape revealed a round configuration in the three groups. Compared to with patients without AS, annular area and sinus of Valsalva area were not different, but patients with tricuspid AS showed smaller STJ areas and longitudinal distances from the aortic annulus to the STJ, resulting in 23% reduction of aortic root volume. In contrast, patients with bicuspid AS had significantly larger annular and Valsalva areas, and larger longitudinal distances between the annulus and the STJ, resulting in 30% increase in aortic root volume. There were weak but significant correlations between aortic annular area and left ventricular volumes (left ventricular end-diastolic volume: r 2 = 0.26, P < .0001; left ventricular end-systolic volume: r 2 = 0.14, P < .0001; stroke volume: r 2 = 0.20, P < .0001).
|Variable||Tricuspid AS||Bicuspid AS||No AS||P|
|Number of patients||57||26||32|
|Age (y)||76 ± 6||66 ± 13 †||71 ± 12||.0003|
|Men||29 (51%)||16 (62%)||16 (50%)||.0062|
|BSA (m 2 )||1.50 ± 0.16||1.57 ± 0.19||1.55 ± 0.20||.2263|
|Systolic BP (mm Hg)||153 ± 25||131 ± 25 †||147 ± 27||.0068|
|Diastolic BP (mm Hg)||81 ± 12||72 ± 16 ∗†||82 ± 16||.0280|
|HTN||42 (74%)||15 (58%)||14 (44%)||.0524|
|Hyperlipidemia||26 (46%)||10 (38%)||6 (19%)||.0473|
|DM||18 (32%)||1 (4%)||4 (13%)||.0934|
|CAD||22 (39%)||3 (12%)||5 (16%)||.0734|
|CRF on HD||18 (32%)||1 (4%)||0 (0%)||.2289|
|HbA 1c (JDS) (%)||5.8 ± 1.0||5.3 ± 0.5||5.7 ± 0.8||.1201|
|LDL (mg/dL)||103 ± 35||104 ± 29||114 ± 39||.4149|
|TG (mg/dL)||116 ± 53||135 ± 94||100 ± 53||.1700|
|Creatinine (mg/dL)||2.9 ± 3.3 ∗||1.1 ± 1.2 †||0.9 ± 0.7||.0002|
|eGFR (mL/min)||41.4 ± 28.8 ∗||67.3 ± 20.3 †||70.7 ± 23.6||<.0001|
|CRP (mg/dL)||0.44 ± 0.66||0.14 ± 0.25||0.16 ± 0.22||.0262|
|Variable||Tricuspid AS||Bicuspid AS||No AS||P|
|LVEDVI (mL/m 2 )||71 ± 22 ∗||74 ± 30 ∗||53 ± 13||.0011|
|LVESVI (mL/m 2 )||34 ± 19 ∗||35 ± 29 ∗||19 ± 9||.0018|
|LVSVI (mL/m 2 )||36 ± 9||39 ± 10||35 ± 8||.2225|
|LVEF (%)||54 ± 13 ∗||58 ± 16 ∗||66 ± 10||.0004|
|LV mass index (g/m 2 )||103 ± 27 ∗||114 ± 26 ∗||82 ± 19||.0012|
|E (cm/sec)||92 ± 37 ∗||74 ± 23 †||70 ± 20||.0020|
|A (cm/sec)||119 ± 36 ∗||73 ± 22 †||80 ± 24||<.0001|
|DT (msec)||282 ± 130 ∗||212 ± 64 †||212 ± 68||.0031|
|E′ (cm/sec)||3.7 ± 1.2 ∗||5.1 ± 1.6 †||5.2 ± 1.5||<.0001|
|E/E′ ratio||28.7 ± 21.3 ∗||15.4 ± 7.9 †||14.3 ± 6.4||.0001|
|AV peak velocity (m/sec)||4.1 ± 0.8||4.2 ± 1.1||—||.7495|
|AV mean PG (mm Hg)||42 ± 17||44 ± 22||—||.5458|
|AVAI 2DCE (cm 2 /m 2 )||0.40 ± 0.12||0.42 ± 0.16||—||.5028|
|AVAI 3D (cm 2 /m 2 )||0.53 ± 0.13 ∗||0.57 ± 0.20 ∗||1.69 ± 0.32||<.0001|
|ELCo (cm 2 /m 2 )||0.46 ± 0.15||0.47 ± 0.20|