Author + information
- Received September 15, 2014
- Revision received January 30, 2015
- Accepted February 17, 2015
- Published online September 1, 2015.
- Hirotsugu Mihara, MD∗,
- Kentaro Shibayama, MD∗,
- Hasan Jilaihawi, MD∗,
- Yuji Itabashi, MD∗,
- Javier Berdejo, MD∗,
- Hiroto Utsunomiya, MD∗,
- Robert J. Siegel, MD∗,†,
- Raj R. Makkar, MD∗ and
- Takahiro Shiota, MD∗,†∗ ()
- ∗Noninvasive Cardiac Laboratory, Cedars-Sinai Heart Institute, Los Angeles, California
- †University of California, Los Angeles, Los Angles, California
- ↵∗Reprint requests and correspondence:
Dr. Takahiro Shiota, Noninvasive Cardiac Laboratory, Cedars-Sinai Heart Institute, 127 South San Vicente Boulevard, A3411, Los Angeles, California 90048.
Objectives The purpose of this study was to determine which echocardiographic parameters, including holodiastolic flow reversal (HDFR) in the descending aorta, were useful for grading of post-procedural aortic regurgitation (PAR) after transcatheter aortic valve replacement (TAVR) using intraprocedural transesophageal echocardiography.
Background Reliable assessment of PAR in a catheterization laboratory is essential for an optimal outcome after TAVR; however, such an assessment has not been determined.
Methods Three hundred eighty patients who underwent TAVR with the Edwards (Irvine, California) balloon-expandable transcatheter heart valve were retrospectively assessed by intraprocedural transesophageal echocardiography. PAR was evaluated by 2-dimensional color Doppler and pulse-wave Doppler in the descending aorta. Using 2-dimensional color Doppler, we measured the cross-sectional area of the vena contracta, the circumferential extent at the aortic annular plane, the longitudinal jet length, and the jet extent (with a mosaic pattern in the left ventricular outflow tract) compared with the location of the tip of the anterior mitral leaflet (AML). Grading of PAR was determined using the following vena contracta cutoffs: mild ≤9 mm2; moderate 10 to 29 mm2; and severe ≥30 mm2. Significant PAR was defined as at least moderate grade.
Results All patients with consistent HDFR had significant PAR. By multivariable analysis, consistent HDFR and the jet extent beyond the tip of AML were independent predictors of significant PAR. Consistent HDFR and jet extent beyond the tip of AML predicted significant PAR with specificities of 100% and 97%, respectively. In contrast, patients with both negative HDFR and a jet extent of less than halfway to the tip of AML had no significant PAR, with 97% specificity.
Conclusions The presence of consistent HDFR and jet extent beyond the tip of AML are indicative of significant PAR after TAVR.
- aortic stenosis
- paravalvular regurgitation
- transcatheter aortic valve replacement
- transesophageal echocardiography
Transcatheter aortic valve replacement (TAVR) is an alternative to surgical aortic valve replacement for inoperable or surgical high-risk patients with severe aortic stenosis (1–3). Post-procedural aortic regurgitation (PAR) is a common complication after TAVR. PAR has been reported in 65% to 96% of TAVR cases (1,4–7). In most cases, the grade of regurgitation is usually only mild. However, moderate and severe grades of PAR occur in 2% to 24% of patients (1,4–7). Multiple studies have demonstrated that moderate and severe grades of PAR are associated with worse clinical outcomes (4,6). Furthermore, the 2-year follow-up of the PARTNER (Placement of Aortic Transcatheter Valves) trial has suggested that even mild paravalvular regurgitation has been associated with increased mortality (3). However, reliable methods to diagnose PAR severity have not been established.
The aortic regurgitation (AR) index is 1 of the valuable prognostic parameters for PAR in a catheterization laboratory (8). However, the echocardiographic quantification of PAR after TAVR is challenging. The recently updated VARC-2 (Valve Academic Research Consortium-2) criteria (9) recommend measuring the circumferential extent (CE) of the color jet at the level of the proximal edge of the transcatheter heart valve (THV) to assess the severity of paravalvular regurgitation. However, this approach has not been validated. Holodiastolic flow reversal (HDFR) in the descending aorta (10–12), as detected by pulse-wave Doppler (PWD), has been used to assess the severity of native chronic AR. However, it has not been validated in patients with TAVR. The aim of this study was to investigate the diagnostic value of several echocardiographic parameters for PAR, including HDFR in the descending aorta, and to propose how these parameters should be used for grading PAR using intraprocedural transesophageal echocardiography (TEE). Furthermore, in the substudy analysis, we compared color Doppler parameters of the short-axis plane of the aortic valve between the level of the aortic annulus and proximal edge of the transcatheter heart valve (proxTHV) to determine which level was the best to evaluate paravalvular regurgitation.
We retrospectively reviewed 400 consecutive severe AS patients who underwent TAVR with a balloon-expandable Edwards Sapien or Sapien XT THV (Edwards Lifesciences, Irvine, California) using intraprocedural TEE from December 2010 to May 2013 in our institute. We excluded 13 patients who underwent the valve-in-valve procedure, 3 patients with past mitral valve replacement using mechanical valves, and 4 patients who did not have TEE images after TAVR. Thus, 380 patients were evaluated. All demographic characteristic data and procedural records were retrieved from medical records. This study was approved by the institutional review board.
Intraprocedural TEE was performed under general anesthesia using an iE33 ultrasound system equipped with a X7-2t TEE ultrasound probe (Philips Medical Systems, Andover, Massachusetts). PAR was assessed by intraprocedural TEE using 2-dimensional (2D) color Doppler and PWD in the descending aorta immediately after THV deployment. If PAR was transvalvular regurgitation, PAR was re-evaluated after removal of the catheter and guidewire. Multiple color Doppler long- and short-axis images of the aortic valve were recorded for PAR assessment with a Nyquist limit of >50 cm/s. Color gain settings were optimized to eliminate random color in areas without flow. The aortic valve was maximally magnified, and then the area of interest was set as small as possible to obtain the highest spatial resolution. For the PWD spectra from the descending aorta, the sample volume was placed to the center of the proximal descending aorta at the periphery of the 90° imaging sector arc, proximally or distally to align as much as possible along the major axis of the aorta (11). The Doppler filter was decreased to its lowest setting to allow detection of low velocities. All images were recorded by experienced echocardiographers with >4 years of experience in the assessment of TAVR echocardiograms.
PAR jet measurement
PAR was assessed by both long- and short-axis imaging of the aortic valve using 2D color Doppler of the intraprocedural TEE. The maximal longitudinal length of a PAR jet with a mosaic pattern from the proxTHV in the left ventricular outflow tract was measured in long-axis imaging (Figure 1A). The qualitative grading for PAR was determined by the maximal longitudinal jet extent with a mosaic pattern (not laminar blue or red color) (Figure 1A) in the left ventricular outflow tract compared with the location of the tip of the anterior mitral leaflet (AML) as follows: grade I: the extent of the jet is less than halfway to the tip of AML; grade II: the extent of the jet is greater than halfway but less than the full distance to the tip of AML; and grade III: the extent of the jet goes beyond the tip of AML (Figure 1A). If multiple PAR jets were observed, the longest jet was used for those measurements.
In the short-axis imaging, the sum of the cross-sectional area of the vena contracta (VCA) was measured at the aortic annular plane (Figure 1B). In the same images, if PAR was paravalvular regurgitation, the circumferential length (Lcirc) and radial length (Lrad) were measured (Figure 1B). The CE of the PAR jet was determined by the percentage of the angle of the vena contracta in the whole THV circumference according to the following formula: angle of the vena contracta ÷ 360 × 100 (Figure 1B). These measurements were performed by paying meticulous attention to identifying an optimal annular plane for the measurement of the short-axis color Doppler parameters of PAR and by using as many videos as possible. In addition, we analyzed multiple 2D color Doppler images from different probe locations with multiple angles and directions in patients with severe acoustic shadowing, and measured the jet parameters in optimal images for each of the regurgitant jets.
PWD measurement of holodiastolic flow reversal in the descending aorta
PWD spectra from the descending aorta were assessed for the presence or absence of HDFR. Diastolic flow was considered to be negative HDFR if flow was predominantly anterograde or if only an early retrograde flow was observed. Positive HDFR was regarded as present if retrograde holodiastolic flow was seen (11). We divided the patients with positive HDFR into 2 groups: a partial positive group if retrograde holodiastolic flow was seen in several beats, but was not seen in all cardiac cycles (Figure 2A); and a consistent positive group if retrograde holodiastolic flow was seen with every cardiac cycle (Figure 2B).
Hemodynamic and angiographic parameters
Simultaneous pressure measurements in the left ventricle and in the ascending aorta were performed with fluid-filled catheters after TAVR. The AR index was calculated according to the following formula: (diastolic blood pressure − left ventricular end-diastolic pressure) ÷ systolic blood pressure × 100) (8). Angiographic assessment of the severity of AR was performed by visual estimation of the concentration of contrast medium in the left ventricle, using the method by Sellers et al. (13).
Grading of PAR
Grading of PAR was determined with 2D color Doppler using intraprocedural TEE. At first, we classified PAR as none or trivial using the visual estimation with 2D color Doppler by 2 independent experienced echocardiographers. The other PAR grades were determined by VCA using the following cutoffs: mild: ≤9 mm2; moderate: 10 to 29 mm2; and severe: ≥30 mm2 in accordance with the cutoff values for effective regurgitant orifice area in the American Society of Echocardiography guideline (14) and VARC-2 criteria (9). If the VCA was not available, we excluded those patients from further analysis. Significant PAR was defined as at least a moderate grade. We investigated the consistency of PAR grading between 2D color Doppler TEE and aortography. All-cause death was investigated from the medical record and compared between the no significant and significant PAR groups.
Substudy analysis for the comparison between the annulus and proxTHV level
From the entire study population, we strictly selected patients with at least mild or more grade paravalvular regurgitation in whom post-procedural short-axis color Doppler images were available exactly at both the level of the aortic annulus and the proxTHV. Simultaneous biplane imaging was used to ensure the optimal selection of these planes using the long-axis view as the primary plane with the orthogonal short-axis view placed at the annulus and the left ventricular outflow tract beneath the skirt of the THV (Figure 3). In those patients, we compared the cross-sectional area of the regurgitant jet, CE, Lcirc, and Lrad between these 2 levels of the short-axis imaging.
Intraobserver and interobserver variability
For intraobserver and interobserver variability, 20 patients were randomly selected and measured by an observer at 2 separate times and also by another independent observer, without knowledge of the results obtained by the other observer.
Data are presented as mean ± SD when the Shapiro-Wilk test showed normal distribution. When nonparametric statistical methods were used, we summarized data with median and quartiles instead of means and SDs. Categorical variables were analyzed as frequency or percentages, as appropriate. Differences between groups with and without significant PAR were assessed by unpaired Student t test or Mann-Whitney U test for continuous variables and by chi-square test or Fisher exact probability test for categorical variables, as appropriate. Differences in the short-axis parameters between the 2 levels of the short-axis in the substudy were assessed by a paired Student t test or a Wilcoxon signed rank test, as appropriate. The associations among parameters for PAR and PAR grading were assessed by Spearman’s correlation coefficient. A kappa statistic was used to determine the agreement on grading of PAR severity between 2D color Doppler TEE and aortography. Using receiver-operating characteristics (ROC) analysis, we investigated which parameters could estimate significant PAR and their cutoff values. The cutoff values were defined using these curves on the basis of the highest sum of the sensitivity and specificity for the prediction of significant PAR. Significant variables in the univariable analysis were included in a multivariable logistic regression analysis with stepwise forward selection method to identify independent variables that predicted significant PAR. Survival curves for time-to-event variables were constructed using Kaplan-Meier estimates and were compared using the log-rank test. To study the effect of significant PAR on mortality, Cox proportional hazards regression was performed. Two-tailed p values <0.05 were considered statistically significant. For color Doppler measurements, the intraobserver and interobserver reproducibility were estimated using 1-way random double-measure and 2-way random single-measure intraclass correlation coefficients. Analyses were performed using SPSS version 19.0 (IBM, Armonk, New York).
Baseline clinical characteristics and TAVR procedure
The data on the baseline clinical characteristics and TAVR procedure are shown in Table 1. PAR grading was determined in 352 patients (93%). PAR after THV deployment was none in 42 patients (12%), trivial in 115 (33%), mild in 100 (28%), moderate in 91 (26%), and severe in 4 (1%). After ballooning a regurgitant THV or a second THV deployment, 91 patients with moderate PAR decreased to mild PAR in 4 patients and remained moderate in 87, and 4 patients with severe PAR decreased to trivial PAR in 2, moderate in 1, and remained severe in 1.
Holodiastolic flow reversal in the descending aorta after TAVR
PWD spectra from the descending aorta were successfully assessed in all of the 128 patients in whom this method was attempted. Three of the 128 patients had PWD spectra in the descending aorta both before and after a repeat ballooning or deployment of a second THV. A total of 131 PWD spectra were evaluated in this study. HDFR was negative in 63 patients, partial positive in 28 patients, and consistently positive in 40 patients. All patients with severe PAR had consistent HDFR. In 5 patients with consistent HDFR and 4 with partial HDFR, grading of PAR was not determined by VCA. In the remaining 35 patients with consistent HDFR, all patients had significant PAR. Twelve (50%) of the remaining 24 patients with partial HDFR had significant PAR (Table 2). Partial HDFR was observed in 2 patients without PAR. The types of HDFR were significantly correlated with jet length, VCA, CE, the AR index (Figure 4), and PAR grading (r = 0.63; p < 0.001).
A representative case of parallel improvement of both VCA and HDFR after a second THV deployment is shown in Figure 5.
Color doppler parameters for PAR
In patients with any grade of PAR, jet extent and VCA were obtained in 80% and 74%, respectively (p = 0.11). Jet length, jet extent, VCA, CE, Lcirc, and Lrad were significantly greater in the significant PAR group than in the group without significant PAR (Table 2). VCA significantly correlated with the grade of jet extent (r = 0.64; p < 0.001) (Figure 6). In the patients with grade I jet extent, 90% of the patients had VCA ≤10 mm2, and all patients had VCA <20 mm2. In contrast, in patients with grade III jet extent, 88% of patients had VCA of ≥10 mm2.
Prediction of significant PAR
ROC curve analyses for predicting significant PAR are listed in Table 3. Using the cutoff values derived from ROC curve analyses, only consistent HDFR and grade III jet extent were independent predictors of significant PAR by multivariable analysis (Table 4). The presence of consistent HDFR and grade III jet extent estimated significant PAR with specificities of 100% and 97%, respectively. In contrast, patients with both negative HDFR and a grade I jet extent had no significant PAR, with 97% specificity (Table 5).
Comparison between echocardiography parameters and other parameters
The AR index was obtained in 92% of patients. The type of HDFR and all color Doppler parameters of the long- and short-axis views at the annulus were significantly correlated with the AR index as follows: type of HDFR: r = –0.26, p = 0.004; jet length: r = –0.22, p = 0.002; jet extent grade: r = –0.21, p = 0.002; VCA: r = –0.21, p = 0.003; CE: r = –0.16, p = 0.03; Lcirc: r = –0.23, p = 0.002; and Lrad: r = –0.16, p = 0.03.
We found 45 patients who had aortography with adequate image quality for grading AR and had echocardiographic PAR grading. In these 45 patients, PAR grading by 2D color Doppler TEE showed modest correlation with aortography (r = 0.50; p < 0.001). PAR grading by 2D color Doppler TEE was inconsistent with that of aortography in 44.4%, was slightly overestimated in 42.2%, and was underestimated in 13.3% compared with aortography (kappa statistic = 0.20; p = 0.01) (Table 6).
The median clinical follow-up was 255 days (interquartile range [IQR]: 71 to 742 days) in the no significant PAR group and 141 days (IQR: 44 to 513 days) in the significant PAR group, and survival data were obtained in all patients. Kaplan-Meier survival estimates suggested that significant PAR was associated with increased late mortality (hazard ratio: 2.4; 95% confidence interval: 1.3 to 4.5; p = 0.004) (Figure 7).
Substudy results for the comparison between the annulus and proxTHV level
Short-axis color Doppler parameters at both the annulus and proxTHV level were measured in 33 patients. In these patients, the ratios of the cross-sectional area of the regurgitant jet, CE, Lcirc, and Lrad at the proxTHV level to the annular level were 2.6 (IQR: 2.0 to 4.0), 1.5 (IQR: 1.0 to 2.0), 1.6 (IQR: 1.0 to 2.0), and 1.8 (IQR: 1.5 to 2.2), respectively, whereas these ratios varied widely depending on the PAR jet directions (Table 7). In this substudy population, VCA and CE at the annulus were significantly correlated with the AR index (r = –0.39, p = 0.03 and r = –0.50, p = 0.01, respectively), whereas the cross-sectional jet area and CE at proxTHV were not correlated (r = –0.09, p = 0.44 and r = –0.21, p = 0.10, respectively).
Reproducibility of measurements
All patients were classified into the same type of HDFR by 2 different observers. Intraclass coefficients of the measurement of the jet length, VCA, and CE were 0.93, 0.96, and 0.88, respectively, for intraobserver variability and 0.94, 0.95, and 0.86, respectively, for interobserver variability.
The main findings of the present study are as follows: 1) the presence of consistent HDFR and grade III jet extent were independent predictors of significant PAR after TAVR by multivariable analysis; and 2) consistent HDFR and grade III jet extent predicted significant PAR with 100% and 97% specificity, respectively.
Assessment of aortic regurgitation after TAVR
Reliable assessment of PAR using echocardiography, especially intraprocedural and intraoperative TEE, is important. VCA, in terms of the effective orifice area, should be a most appropriate parameter for assessing PAR grading, as was previously validated by Gonçalves et al. (15), who used 3-dimensional (3D) transthoracic echocardiography color Doppler, by Altiok et al. (16) who used cardiac magnetic resonance in patients with TAVR, and by Fang et al. (17) in patients with native aortic valve regurgitation. Khalique et al. (18) used 3D TEE color Doppler to determine the effective regurgitant orifice area in patients with paravalvular AR after TAVR. However, the image qualities and spatial resolution of 3D transthoracic echocardiography and 3D TEE color Doppler are lower than those of 2D TEE. In addition, this 3D color method can be time-consuming, and thus its application during the TAVR procedure in a cardiac catheterization laboratory is limited. Consequently, we used 2D TEE VCA as a reference for grading PAR. However, this VCA, the type of HDFR, and all other color Doppler parameters were significantly correlated with the AR index.
For assessing paravalvular AR, VARC-2 criteria (9) recommend measuring the CE of the regurgitant jet at the proxTHV level. However, for the surgical prostheses, the American Society of Echocardiography guideline (14) recommends measuring the jet at the level of the prosthesis sewing ring, which corresponds to the annular level. However, neither of these methods has been validated. Hahn et al. (19) assessed paravalvular AR by measuring the cross-sectional area of the regurgitant jet at the proxTHV level using 2D TEE. From our substudy analysis, the ratio of the regurgitant jet area from the annulus to the proxTHV level varied widely, from 1.1 to 9.4. Furthermore, the parameters at the proxTHV level were not correlated with the AR index. Therefore, our findings suggest that the cross-sectional area of the PAR jet at the proxTHV level may not be ideal for the quantification of PAR.
HDFR in the descending aorta
In our study, in addition to the high diagnostic value of consistent HDFR for significant PAR, PWD in the descending aorta was assessed in 100% of patients as opposed to the VCA in 74% of patients. Furthermore, its evaluation was highly reproducible. Therefore, it is a useful parameter for PAR evaluation. In native valve chronic AR assessment, diastolic flow reversal in the descending aorta has been used to determine severity of AR by echocardiography (10–12) and by cardiac magnetic resonance (20). In most of these previous studies, transthoracic echocardiography was used, which has a limited window from the suprasternal notch, and sometimes has significant artifacts, leading to the inability to reliably detect end-diastolic reversal in the descending aorta. In our study, we used intraoperative TEE, which has a greater Doppler angle with the descending flow. However, Sutton et al. (11) obtained adequate spectra from the proximal descending aorta using TEE in >95% of patients, which is in line with our results.
A reliable intraprocedural echocardiographic grading of PAR is important for optimizing the TAVR result. VCA can be used as a reference for grading PAR. However, this method is cumbersome and needs experienced echocardiographers for both imaging and analyzing due to the through-plane phenomenon in the short-axis view of the aortic valve, if this measurement was strictly done at the exact annular level. We found that VCA could be obtained in only 74% of patients with PAR, whereas longitudinal jet extent was reproducibly measured in 80%, and PWD spectra in the descending aorta were obtained in 100% of patients. Therefore, in addition to VCA assessment, we proposed here to assess HDFR in the descending aorta and “classical” longitudinal jet extent for grading of PAR. These methods are easy to use, highly reproducible, and convenient intraprocedural parameters for PAR. Therefore, they can serve as surrogates or alternatives when direct visualization of the PAR orifice is not optimal due to the artifact.
First, this was a retrospective study. However, echocardiographic data were recorded by one of the most experienced echocardiography teams on TAVR. Second, the assessment of the descending aorta flow by TEE had several limitations, such as recording during hemodynamic instability, a large Doppler interrogation angle with the direction of the blood stream, reduced aortic compliance, and increased heart rates. Nonetheless, the presence of HDFR successfully predicted significant PAR. Third, this VCA method for grading PAR by 2D color Doppler TEE was not strictly validated. However, this VCA method by 3D transthoracic echocardiography was validated by other investigators (15,16). In addition, this VCA method by 3D color Doppler TEE was used for grading paravalvular regurgitation after TAVR in recent studies (18,19). Fourth, eccentric or curved jets might be underestimated by the longitudinal jet length and extent or overestimated by the VCA. Therefore, the direction of the regurgitant jet should be taken into account for grading PAR using these jet parameters. Fifth, we used VCA as a reference method. Therefore, short-axis parameters such as CE, Lcirc, and Lrad, which are part of the components of VCA, should be excluded from the multivariable analysis due to their mutual high correlation.
The presence of consistent HDFR and jet extent beyond the tip of AML are indicative of significant PAR after TAVR.
COMPETENCY IN MEDICAL KNOWLEDGE: PAR after TAVR is a relatively common complication and is associated with an increased risk for cardiovascular mortality after TAVR. Reliable grading of PAR in the catheterization laboratory is essential for an optimal outcome after TAVR. This study explores the utility of easily applicable semiquantitative parameters of aortic regurgitation for grading PAR during TAVR.
TRANSLATIONAL OUTLOOK: Techniques for assessing PAR after TAVR require further standardization in multicenter settings, particularly for determining the threshold of PAR in which additional intervention in the catheterization or hybrid laboratory leads to improvement in clinical outcomes following TAVR.
The authors thank Dr. and Mrs. Paul I. Terasaki for their kind support and encouragement.
Dr. Jilaihawi is a consultant for Edwards Lifesciences, St. Jude Medical, and Venus Medtech. Dr. Siegel is a consultant for Abbott; and is a member of the Speakers Bureau for Philips Ultrasound. Dr. Makkar has received research grants from Cordis, Edwards Lifesciences, Medtronic, Abbott Vascular, Capricor, and St. Jude Medical; is a proctor for Edwards; and is a consultant for Medtronic. Dr. Shiota is a speaker for Philips Ultrasound. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- anterior mitral leaflet
- aortic regurgitation
- circumferential extent
- holodiastolic flow reversal
- circumferential length
- radial length
- post-procedural aortic regurgitation
- proximal edge of the transcatheter heart valve
- pulse-waved Doppler
- receiver-operating characteristic
- transcatheter aortic valve replacement
- transesophageal echocardiography
- transcatheter heart valve
- cross-sectional area of the vena contracta
- Received September 15, 2014.
- Revision received January 30, 2015.
- Accepted February 17, 2015.
- American College of Cardiology Foundation
- Athappan G.,
- Patvardhan E.,
- Tuzcu E.M.,
- et al.
- Detaint D.,
- Lepage L.,
- Himbert D.,
- et al.
- Sinning J.M.,
- Vasa-Nicotera M.,
- Chin D.,
- et al.
- Sinning J.M.,
- Hammerstingl C.,
- Vasa-Nicotera M.,
- et al.
- Tribouilloy C.,
- Avinee P.,
- Shen W.F.,
- Rey J.L.,
- Slama M.,
- Lesbre J.P.
- Zoghbi W.A.,
- Chambers J.B.,
- Dumesnil J.G.,
- et al.
- Khalique O.K.,
- Kodali S.K.,
- Paradis J.M.,
- et al.