Background
Successful transcatheter aortic valve implantation (TAVI) mandates comprehensive, accurate multimodality imaging. Echocardiography is involved at all key stages and, with the advent of real-time three-dimensional (3D) transesophageal echocardiography, is uniquely placed to enable periprocedural monitoring. The investigators describe a comprehensive two-dimensional (2D) and 3D echocardiographic protocol, and the additional benefits of 3D TEE, within a high-volume TAVI program.
Methods
TAVI was performed with 2D and 3D transesophageal echocardiographic and fluoroscopic guidance in consecutive high-risk patients with symptomatic severe aortic stenosis. The role of TEE, including the additive value of 3D TEE, was examined, and procedural and echocardiographic outcomes were evaluated. A 3D sizing transcatheter heart valve (THV) strategy was used, except as mandated by study protocol.
Results
Procedural success was achieved in 99% of 256 patients (mean age, 82.9 ± 7.1 years, mean logistic European System for Cardiac Operative Risk Evaluation score, 21.6 ± 11.2%; mean aortic valve area, 0.63 ± 0.19 cm 2 ), with no procedural deaths. Acceptable 2D and 3D transesophageal echocardiographic images were achieved in all patients. Aortic valve annular dimensions by 2D transthoracic echocardiography, 2D TEE, and 3D TEE were 21.6 ± 1.9 mm, 22.5 ± 2.2 mm ( P < .001), and 23.0 ± 2.0 mm ( P = .004 vs 2D TEE), respectively. The 2D THV sizing strategy would have changed THV selection in 23% of patients, downsizing in most. Three-dimensional TEE provided superior spatial visualization and anatomic orientation and optimized procedural performance. Postprocedural mild, moderate, and severe paravalvular aortic regurgitation was observed in 24%, 3%, and 0% of patients, respectively, with no or trace transvalvular aortic regurgitation in 95%. A second valve was successfully deployed in five patients, and TEE detected five other periprocedural complications.
Conclusions
A systematic, comprehensive echocardiographic protocol, incorporating the additional benefits of 3D TEE, has a vital role within a TAVI program and, combined with a 3D THV sizing strategy, contributes to excellent outcomes.
Transcatheter aortic valve (AV) implantation (TAVI) has emerged as a potential therapeutic option for patients with symptomatic severe aortic stenosis (AS) deemed at high risk or unsuitable for conventional surgical AV replacement. Although procedural success rates are high, with promising short-term and medium-term outcomes, this should be balanced against 30-day major stroke, major vascular complication, and mortality rates of 2% to 5%, 3% to 16%, and 5% to 19%, respectively. Typical patients undergoing TAVI are the frail elderly with multiple medical comorbidities and potential clinical instability, and as such, appropriate case selection and optimum perioperative management require a truly multidisciplinary approach. Comprehensive, accurate multimodality imaging, potentially involving echocardiography, angiography and fluoroscopy, computed tomographic angiography, and cardiac magnetic resonance imaging techniques, and the involvement of appropriately trained imaging specialists, is mandated within a successful TAVI program. This enables complete patient assessment, appropriate case selection, and procedural guidance, including the choice of access route and prosthesis type and size, in addition to the detection of complications.
Although the differing imaging modalities should be seen as complementary techniques, echocardiography plays a pivotal role at all stages within a TAVI program and, at present, is uniquely placed to enable real-time procedural guidance and detection of complications. Furthermore, the emergence of real time three-dimensional (3D) transesophageal echocardiography (TEE), with images that can be readily manipulated online to display cardiac structures or enable quantitative analysis, has been a major advance. In particular, 3D TEE with multiplanar reconstruction provides more detailed cross-sectional assessment of AV annular anatomy and, as such, may optimize prosthesis selection and potentially outcome data. However, we believe it is important to evaluate factors other than simply AV annular dimension when selecting prosthesis size, particularly in borderline patients in whom the relative risks of implantation of a smaller or larger prosthesis should be considered. These factors include AV annular dimension relative to left ventricular outflow tract and aortic root measurements; the severity, eccentricity, and volume of AV leaflet calcification and whether there is sufficient space in the sinuses to accommodate this; the position of the coronary ostia relative to the AV annulus and any calcification; and the morphology of the left ventricular outflow tract and aortomitral continuity (presence of calcification, basal septal hypertrophy). Two-dimensional (2D) and 3D TEE are complementary techniques that are well placed to evaluate these multiple factors important in determining optimal prosthesis selection.
The aims of this study were to describe a comprehensive 2D and 3D echocardiographic protocol, and our experience of the additional benefits of 3D TEE, within a successful, high-volume multidisciplinary TAVI program.
Methods
Patients
All consecutive patients with symptomatic severe AS undergoing TAVI at King’s Health Partners (KHP) between August 2007 and October 2010 were entered prospectively in a local registry and included in the present study. TAVI is approved for use in patients with symptomatic severe AS deemed unfit for conventional surgical AV replacement (logistic European System for Cardiac Operative Risk Evaluation score > 20, Society of Thoracic Surgeons score > 10%, or turned down by two separate cardiothoracic surgeons) whose life expectancy is not severely limited. Potential TAVI candidates referred to the KHP program undergo detailed and systematic evaluation for suitability that includes multidisciplinary team review, echocardiography (transthoracic echocardiography [TTE] and 2D and 3D TEE), coronary angiography, right-heart catheterization, aortofemoral and ileofemoral angiography, and computed tomography (CT). All patients provided informed written consent to the procedures.
TAVI
TAVI procedures, including transapical cases, were performed in an adapted cardiac catheter laboratory using the Edwards SAPIEN transcatheter heart valve (THV) (Edwards Lifesciences, Irvine, CA) in all patients by a dedicated team of interventional cardiologists, cardiothoracic surgeons, cardiac imaging specialists, cardiac anesthesiologists, and specialist nursing staff members, using techniques described extensively elsewhere. Choice of access approach (transapical, transfemoral, or transaortic) was made by the multidisciplinary team, taking into consideration cardiac and vascular anatomy and patient comorbidities.
Echocardiography
All patients referred to the TAVI program underwent initial TTE to confirm the diagnosis of severe AS and to assess anatomic suitability. If there was any uncertainty after TTE, TEE was performed as part of the workup; more recently, however, preprocedural 2D and 3D TEE has been performed in all patients. All TAVI procedures were performed with complementary TEE (2D and 3D) and fluoroscopic or angiographic guidance.
TTE and TEE were performed using an iE33 xMATRIX echocardiographic system with X3-1 xMATRIX Array and X7-2t Live 3D transducers, respectively (Philips Medical Systems, Best, The Netherlands). This system enables real-time 3D TEE imaging (Live 3D and Live 3D zoom modes), as well as simultaneous imaging in two orthogonal planes (Live xPlane), triggered 3D full-volume and 3D color acquisitions, and conventional 2D multiplane transesophageal echocardiographic, color, and spectral Doppler imaging. In the full-volume modes, four-beat acquisitions were typically used.
A schematic representation of the KHP TAVI echocardiography protocol is presented in Figure 1 , and key steps are described in greater detail. Accurate determination of AV annular dimension was performed using a combination of 2D and 3D techniques. The aortic root was visualized on 2D TEE in the midesophageal long-axis view. The digital zoomed image was frozen in midsystole with the AV clearly visualized. The AV annular dimension was measured from the hinge point of the noncoronary cusp with the posterior aortic wall to the hinge point of the right coronary cusp and the anterior aortic wall, perpendicular to the long axis of the root. Manipulation of the probe may be required to avoid poor visualization because of acoustic shadowing from calcified deposits. A 3D zoom volume was acquired and multiplanar reconstruction performed online using the manufacturer’s software (QLAB-3DQ; Philips Medical Systems) to facilitate measurement of the anatomically correct anteroposterior dimension, as well as maximum and minimum AV annular dimensions ( Figure 2 and Video 1 ; available at www.onlinejase.com ). A 3D THV sizing strategy was used, except in patients included in the Placement of Transcatheter Aortic Valves Europe study, who were sized according to 2D AV annular dimension as mandated by study protocol, although 3D data were also recorded. The average of the 3D maximum and minimum annular dimensions was used for sizing. Multiplanar reconstruction from a 3D volume was also used to assess AV annular–coronary ostial height, when required ( Figure 3 ). In addition to standard 2D transesophageal echocardiographic techniques and recommendations for the assessment of the etiology and severity of AR, where appropriate and valid, 3D zoom and full-volume transesophageal echocardiographic color modes were also used ( Video 2 ; available at www.onlinejase.com ). AV gradients were recorded from the 2D transesophageal echocardiographic transgastric and deep transgastric views, with Doppler alignment optimized where possible.
Standard 2D and complementary 3D transesophageal echocardiographic techniques were used for intraprocedural guidance, in addition to fluoroscopy. When the approach to or crossing of the stenosed AV was difficult, manipulation of a Live 3D volume to show the guidewire and AV en face or, alternatively, demonstration of orthogonal views using xPlane was used. Three-dimensional transesophageal echocardiographic live imaging was used to visualize and facilitate guidewire, delivery catheter, and balloon introduction, positioning, and retrieval ( Figure 4 A). Three-dimensional transesophageal echocardiographic live xPlane imaging, providing orthogonal views, was used during balloon aortic valvuloplasty with long loop acquisition to enable immediate review and evaluation of sizing, behavior of the native AV leaflets and calcified material during inflation, as well as confirming balloon stability and optimal inflation and deflation ( Figure 4 B). Positioning and deployment of the THV was guided using 3D transesophageal echocardiographic live imaging, again with long loop acquisition ( Figure 5 ). Comprehensive transesophageal echocardiographic evaluation was performed after THV deployment to assess procedural success and to enable the early detection of complications.
Outcomes
Procedural success was defined as the implantation of a functional AV prosthesis without intraprocedural mortality. All major procedural complications were recorded. Post-TAVI AV hemodynamic status was assessed using standard echocardiographic techniques and criteria, and all procedural complications detected by echocardiography were documented.
Statistical Analysis
All data were collected prospectively. Continuous variables are presented as mean ± SD unless stated otherwise. Categorical variables are presented as percentages and frequencies. Group comparisons were analyzed using Student’s t tests for continuous variables and Fisher’s exact tests or χ 2 tests for categorical variables. Paired t tests were used to compare continuous variables before and after intervention. Differences were considered statistically significant at P < .05.
Intraobserver and interobserver agreement for transesophageal echocardiographic AV annular dimensions was assessed in a smaller cohort of 20 patients. Two expert readers, able to select the best cardiac cycle and image at each reading and blinded to previous measurements, measured 2D and 3D annular dimensions in each patient; one reader repeated the measurements on the same patient cohort on a separate occasion >1 week later. Data are presented as intraclass correlation coefficients and as absolute differences between repeated measurements expressed as percentages of their mean values.
Results
This study included 256 patients with symptomatic severe AS who underwent TAVI at our institution. Baseline clinical and echocardiographic characteristics are presented in Tables 1 and 2 , respectively. Patients were typically >75 years of age (86%; mean age, 82.9 years) with significant comorbidities and marked symptoms (189 [74%] with New York Heart Association class III and IV symptoms). All had echocardiographic data consistent with severe AS (mean peak and mean pressure gradients, 78.2 ± 25.0 and 47.1 ± 1.0 mm Hg, respectively; mean AV area, 0.63 ± 0.19 cm 2 ), with concomitant mitral regurgitation and aortic regurgitation (AR) moderate or greater in 22% and 15% of patients, respectively. Most patients had normal left ventricular systolic function, with only 16 (6%) with severe left ventricular systolic dysfunction, defined as left ventricular ejection fraction < 30%. Logistic European System for Cardiac Operative Risk Evaluation scores were higher in the transapical or transaortic group compared with the transfemoral group, and in addition, these patients had a higher incidence of peripheral vascular disease and severe chronic kidney disease. There were no differences in baseline echocardiographic characteristics between the transapical or transaortic and transfemoral groups.
| Variable | All ( n = 256) | Transapical/transaortic ( n = 143) | Transfemoral ( n = 113) | P |
|---|---|---|---|---|
| Age (y) | 82.9 ± 7.1 | 82.6 ± 7.2 | 83.1 ± 7.1 | .72 |
| Women | 138 (53.9%) | 75 (52.5%) | 63 (55.8%) | .60 |
| BMI (kg/m 2 ) | 26.5 ± 5.7 | 25.6 ± 5.0 | 27.5 ± 6.3 | .99 |
| Logistic EuroSCORE (%) | 21.6 ± 11.2 | 22.9 ± 12.0 | 20.2 ± 9.1 | .04 |
| Cardiac rhythm | ||||
| SR | 170 (66.7%) | 90 (63.4%) | 80 (71.0%) | .23 |
| AF | 70 (27.5%) | 41 (28.9%) | 29 (25.7%) | .78 |
| PPM | 22 (8.6%) | 16 (11.3%) | 6 (5.3%) | .12 |
| Diabetes | 63 (24.7%) | 31 (21.8%) | 32 (28.3%) | .23 |
| Hypertension | 171 (67.1%) | 97 (68.3%) | 74 (65.5%) | .64 |
| Cerebrovascular disease | 38 (14.9%) | 20 (14.1%) | 18 (15.9%) | .68 |
| Peripheral vascular disease | 56 (22.0%) | 48 (33.8%) | 8 (7.1%) | <.0001 |
| Chronic obstructive pulmonary disease | 54 (21.1%) | 32 (22.4%) | 22 (19.5%) | .64 |
| Prior cardiac surgery | 57 (22.3%) | 34 (23.8%) | 23 (20.4%) | .55 |
| Prior coronary angioplasty | 38 (14.8%) | 23 (16.2%) | 15 (13.3%) | .84 |
| Porcelain aorta | 32 (12.5%) | 20 (14.1%) | 12 (10.6%) | .40 |
| Coronary artery stenosis >50% | 118 (46.0%) | 74 (51.8%) | 44 (38.9%) | .12 |
| Carotid artery stenosis >50% | 43 (17.1%) | 28 (19.9%) | 15 (13.6%) | .19 |
| Creatinine (μmol/L) | 112.9 ± 64.7 | 117.8 ± 71.7 | 106.8 ± 54.2 | .08 |
| eGFR (mL/min/1.73 m 2 ) | 57.0 ± 20.4 | 55.0 ± 20.9 | 59.6 ± 19.5 | .09 |
| eGFR < 60 mL/min/1.73 m 2 | 139 (58.1%) | 88 (66.7%) | 52 (51.0%) | .047 |
| Variable | All ( n = 256) | Transapical/transaortic ( n = 143) | Transfemoral ( n = 113) | P |
|---|---|---|---|---|
| LVEF | ||||
| Good (≥50%) | 67% | 67% | 66% | 1.00 |
| Fair (30%–49%) | 27% | 26% | 29% | .33 |
| Poor (<30%) | 6% | 7% | 4% | .43 |
| Peak gradient (mm Hg) | 78.2 ± 25.0 | 75.8 ± 25.1 | 81.1 ± 24.8 | .95 |
| Mean gradient (mm Hg) | 47.1 ± 16.3 | 45.6 ± 15.8 | 49.2 ± 17.6 | .10 |
| AV area (cm ) | 0.63 ± 0.19 | 0.64 ± 0.18 | 0.62 ± 0.19 | .22 |
| Moderate or greater mitral regurgitation | 22% | 16% | 12% | .49 |
| Moderate or greater AR | 15% | 13% | 4% | .06 |
| Aortic annular dimension on TTE (mm) | 21.6 ± 1.9 | 21.6 ± 2.0 | 21.6 ± 1.8 | .81 |
| Aortic annular dimension on TEE (mm) | 22.2 ± 2.1 | 22.2 ± 2.3 | 22.2 ± 1.8 | .99 |
| AV calcification severity | ||||
| Mild | 1% | 2% | 1% | 1.00 |
| Moderate | 48% | 53% | 41% | .09 |
| Severe | 51% | 45% | 58% | .07 |
Acceptable 2D and 3D transesophageal echocardiographic images, as per the TAVI echocardiographic protocol used at KHP ( Figure 1 ), were achieved in all patients. AV annular dimensions by 2D TTE, 2D TEE, and 3D TEE were 21.6 ± 1.9 mm, 22.5 ± 2.2 mm ( P < .001), and 23.0 ± 2.0 mm ( P = .004 vs 2D TEE), respectively. Intraobserver and interobserver agreement for annular dimension was good for both 2D and 3D techniques (intraclass correlation coefficients, 0.83 and 0.75 for 2D and 3D intraobserver agreement and 0.82 and 0.77 for 2D and 3D interobserver agreement; absolute difference between repeated measurements as a percentage of mean value, 4.4 ± 3.3% and 5.4 ± 4.5% for 2D and 3D intraobserver agreement and 5.6 ± 3.0% and 5.0 ± 3.2% for 2D and 3D interobserver agreement; P > .05). The 23-mm, 26-mm, and 29-mm THVs were implanted in 43%, 55%, and 2% of patients, respectively, through transapical (52%), transfemoral (44%), or transaortic (4%) approaches. THV sizing strategies, based solely on 2D annular dimension, would have changed in 23% of patients, with selection of smaller prostheses in most (19%). Procedural success was achieved in 99% of patients, and there were no procedural deaths. The procedure was abandoned after balloon aortic valvuloplasty in one patient with an annular dimension of 19 mm and basal septal hypertrophy; inflation of a 20-mm balloon showed inadequate space, particularly in the left ventricular outflow tract, to safely accommodate a 23-mm THV, in addition to concerns regarding potential obstruction of the right coronary ostium by a calcified native leaflet ( Figure 4 B).
Immediate postprocedural echocardiographic data are shown in Table 3 . Peak and mean gradients fell from 78.2 ± 25.0 and 47.1 ± 16.3 mm Hg to 10.5 ± 5.8 and 5.3 ± 3.5 mm Hg, respectively ( P < .005). Immediately after THV deployment, 73% of patients had either no or trace AR and 24% had mild paravalvular AR; only 3% had moderate, and none severe, paravalvular AR as assessed by TEE ( Figure 6 ). Three-dimensional cover index, defined as 100 × (implanted THV size − 3D AV annular dimension)/implanted THV size, for patients with no or trace paravalvular AR was 8.4 ± 6.7% and for patients with mild or moderate paravalvular AR was 3.8 ± 6.4% ( P < .001). Patients with no or trace paravalvular AR did not differ from those with mild or moderate paravalvular AR, with respect to the discrepancy between 3D and 2D transesophageal echocardiographic annular dimensions (0.56 ± 1.3 and 0.58 ± 1.6 mm, respectively, P = .94).
