Author + information
- Brandon M. Wiley, MD,
- Jason C. Kovacic, MD, PhD,
- Sandeep Basnet, MD,
- Amaki Makoto, MD,
- Farooq A. Chaudhry, MD,
- Annapoorna S. Kini, MD,
- Samin K. Sharma, MD and
- Partho P. Sengupta, MD∗ ()
- ↵∗Zena and Michael A. Wiener Cardiovascular Institute, Mount Sinai Medical Center, One Gustave L. Levy Place, New York, New York 10029
The success of transcatheter aortic valve replacement (TAVR) depends on the appropriate evaluation of the aortic annulus because significant errors in sizing can lead to post-procedure paravalvular leak, prosthesis migration, coronary artery occlusion, or annulus rupture. In a previously published review, Kasel et al. (1) proposed a novel method called the “turnaround rule” as a technique to improve annular measurements by 3-dimensional (3D) transesophageal echocardiography (TEE). We present data validating the use of this method for intraoperative, real-time measurement of annular dimensions that can be used to confirm values obtained from multidetector computed tomography (MDCT) and thus ensure appropriate sizing of the aortic prosthesis.
From September 1, 2012 to July 31, 2014, 141 consecutive patients with severe aortic stenosis underwent TAVR at our institution and were considered for this study. Of this patient population, 74 had both pre-procedure MDCT and intraprocedural 3D-TEE. (Excluded patients had 2-dimensional transthoracic echocardiogram at the time of the procedure.) Within this cohort, 4 patients had bioprosthetic valves and were excluded from this study. Therefore, the final study cohort included 70 patients. Fifty-five patients received SAPIEN 9000TFX (Edwards Lifesciences, Irvine, California) and 15 patients received CoreValve (Medtronic, Inc., Minneapolis, Minnesota) valves. The patients were nonconsecutive because patients who did not undergo both imaging modalities were excluded. No patients were excluded due to poor image quality.
TEE was performed and processed with a commercially available 3D software package (Vivid, GE Healthcare, Fairfield, Connecticut). The aortic valve annulus was measured using the “turnaround rule” technique as described by Kasel et al. (1) (Figure 1). With this technique, 3D-TEE is performed over zoom mode to obtain a loop containing the entire aortic root, the left ventricular outflow tract, and part of the ascending aorta. With the short-axis view of the aortic valve in the transverse plane as the reference point, the orthogonal sagittal and coronal planes are aligned parallel to the long axis of the ascending aorta. Rotation of the sagittal and coronal planes identifies the most caudal hinge points of the aortic cusps. The level of the hinge points in the transverse plane defines the virtual aortic annulus. The average diameter (Darea) was calculated from the annular area (Darea = 2 × √[area/π]). The mean diameter (Dmean) was calculated from the average of the maximum and minimum diameters. Eccentricity index was calculated as the ratio of maximal diameter over minimal diameter and is a measure of the elliptical nature of the valve.
MDCT imaging was acquired using a 256-slice MDCT imager (Brilliance iCT, Philips Healthcare, Andover, Massachusetts) or 64-slice MDCT imager (LightSpeed VCT, GE Healthcare). MDCT images were analyzed independently from the 3D-TEE and by different personnel, using either Philips Intellispace Portal or Osirix MD version 1.4.2 software. All measurements presented in this study were acquired for the specific purpose of clinical TAVR sizing and implantation and, therefore, represent “real-world” data.
The mean annular area was 40.45 ± 7.71 mm2 by 3D-TEE and 42.11 ± 7.51 mm2 by MDCT (mean difference 1.66 ± 4.32 mm2; p = 0.002). The mean perimeter was 71.45 ± 6.84 mm by 3D-TEE and 74.42 ± 6.52 mm by MDCT (mean difference 2.97 ± 4.29 mm; p < 0.001). The Dmean was 22.86 ± 2.22 mm by 3D-TEE and 23.52 ± 2.20 mm by MDCT (mean difference 0.68 ± 1.60 mm; p = 0.003). The Darea was 22.59 ± 2.15 mm by 3D-TEE and 23.07 ± 2.03 mm by MDCT (mean difference 0.47 ± 1.21 mm; p = 0.002).
The Pearson coefficient correlations (Darea r = 0.8337, p < 0.0001; Dmean r = 0.7396; p < 0.0001; annular area r = 0.8392, p < 0.0001; perimeter r = 0.7949, p < 0.0001) showed a strong correlation between 3D-TEE and MDCT measurements. Bland-Altman plots showed that 3D-TEE mean values were smaller than those of MDCT. The plots had the following mean differences and 95% limits of agreement: Darea −0.47 mm (−2.84 and 1.89 mm), Dmean −0.67 mm (−3.80 and 2.50 mm), annular area −1.66 mm2 (−10.13 and 6.80 mm2), and perimeter −2.97 mm (−11.38 and 5.43 mm).
The eccentricity index was not significantly different between the 2 methods (3D-TEE 1.21 ± 0.13; MDCT 1.20 ± 0.10; p = 0.49). The intraclass correlation coefficient for interobserver variability in 3D-TEE annular area measurement was 0.91 (95% confidence interval: 0.77 to 0.96) and for intraobserver variability between MDCT and 3D-TEE was 0.81 (95% confidence interval: 0.69 to 0.89).
These results parallel those of prior studies, which showed that 3D-TEE aortic annular measurements were smaller than those derived from MDCT (2). Although statistically significant, the mean differences found in this study were small and were not clinically relevant. In comparison to prior published data, the incorporation of the “turnaround rule” in aortic annular measurements improved the correlation between 3D-TEE and MDCT values and strengthened interobserver agreement in 3D-TEE measurements (2). These results are similar to those of a recent study by Khalique et al. (3) that also used a technique of cross-sectional 3D-TEE modeling (Q-lab MVQ software, Philips Healthcare).
This study validates the “turnaround rule” as a robust technique that significantly improves the correlation of 3D-TEE and MDCT measurements, reduces interobserver variability, and accurately determines the virtual aortic annulus size. The “turnaround rule” can be incorporated into intraprocedural TEE protocols for rapid confirmation of annulus dimensions and prosthesis sizing during TAVR procedures.
Please note: Dr. Sharma serves on the Speakers Bureaus of Abbott Vascular, AngioScore, Boston Scientific, Cardiovascular Systems Inc., and Lilly/Daiichi-Sankyo. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose. Sherif Nagueh, MD, served as Guest Editor for this paper.
- American College of Cardiology Foundation
- Kasel A.M.,
- Cassese S.,
- Bleiziffer S.,
- et al.
- Jilaihawi H.,
- Doctor N.,
- Kashif M.,
- et al.
- Khalique O.K.,
- Kodali S.K.,
- Paradis J.M.,
- et al.