Author + information
- Published online January 7, 2019.
- Linda D. Gillam, MD, MPH∗ ( and )
- Leo Marcoff, MD
- Department of Cardiovascular Medicine, Morristown Medical Center/Atlantic Health System, Morristown, New Jersey
- ↵∗Address for correspondence:
Dr. Linda D. Gillam, Cardiac Administration, Gagnon Cardiovascular Institute at Morristown Medical Center, 100 Madison Avenue, Morristown, New Jersey 07960.
Transcatheter aortic valve replacement (TAVR) has proven to be a disruptive technology. As trials have provided support for its use as an alternative to surgical aortic valve replacement in groups at progressively lower surgical risk, it is estimated that by 2025, TAVR will account for roughly 75% of aortic valve replacements, assuming that transcatheter aortic valve (TAV) and surgical aortic valve (SAV) durability are similar. Thus, the ability to identify abnormal TAV function is important, with echocardiography being the primary technique used for this purpose. In this issue of iJACC, Hahn et al. (1) provide important tools by providing valve and size specific normal values for mean gradients, effective orifice area (EOA), and Doppler velocity index (DVI) as well as instructions as to how these assessments should be performed. There have been prior reports of normal values from single center and registry experiences as well as a recent meta-analysis (2) that included data from >27,000 valves but with largely site-measured values. Although there have been trial- and device-specific reports that have included core laboratory-adjudicated results, this is the first paper that has provided core laboratory-adjudicated values for both balloon expandable and self-expanding valves, including multiple valve generations. Additionally, this is the first study to provide guidance as to the echocardiographic methods for assessing TAV function and to explore the interaction between native annular anatomy and TAVs, which unlike SAVs, do not have a rigid ring.
Importance of Core Laboratory-Adjudicated Results and a Standardized Approach to Image Acquisition
Although site-adjudicated results have played a role in TAVR evaluation, variable expertise in and approaches to image acquisition and interpretation and the possibility of underappreciated bias provide strong arguments for the superiority of core laboratory-adjudicated results for defining normal function (3). The impartiality, standardization of measurement methods, and attention to intraobserver and interobserver variability inherent in core laboratory measurements are important differentiators.
Typically core laboratories also set standards for image acquisition, as is the case with the studies included in this report. Particularly important for TAVs is the recognition of flow characteristics that differ from those of SAVs with flow acceleration at the inlet to the valve stent and again at the level of the leaflets. Thus, this paper’s detailed description of how echo-Doppler evaluation of TAVs should be performed is arguably as important a contribution to the field as the reference normal values. The authors emphasize that matching left ventricular outflow tract (LVOT) diameter measurement with pulsed wave (PW) Doppler sample volume position, a core principle of echo-Doppler flow calculations, translates to measuring the outer edge–to–outer edge diameter of the lower end of the stent with precise positioning of the sample volume in the same location, that is, just proximal to the stent inlet (4,5). Only when these measurements are challenging, typically with a low-positioned self-expanding valve, do the authors recommend measuring both the diameter (inner edge to inner edge) and PW spectrum within the stent just proximal to the leaflets. Without close attention to these concepts and precise PW sample volume placement, inaccurate measurements of stroke volume and consequently EOA will occur. Particularly concerning is the impact of inconsistent PW sampling and mismatching stent/PW measurements. For example, measuring the diameter at the lower edge of the stent but the PW spectrum just inside the stent where flow acceleration has already occurred results in inappropriately high LVOT stroke volume and EOA calculations. Inconsistency over follow-up may incorrectly suggest that valve function has changed when, in fact, it is stable.
It would also have been helpful if the authors shared their approach to measuring the PW spectra when spectral broadening, which renders the modal velocity indistinct, is unavoidable as in Figure 1C (1). Did they trace the outer envelope, alter the brightness of the display in hopes of teasing out the modal velocity, or use some other approach?
Relation Between Valve Size and Hemodynamics
It seems intuitive that there should be a relation between EOA and nominal valve size, as has been shown so consistently with surgical valves. This in fact was the case. However, note that for TAVs, unlike SAVs, the size of the stent directly has an impact on the EOA calculation because it defines the LVOT diameter according to the convention used here. DVI, a surrogate for EOA, is typically used when there is uncertainty as to LVOT diameter because the calculation requires Doppler spectra only. The impact of valve size on DVI was variable but even where there were statistically significant differences, the absolute differences tended to be small and often within the measurement error of the technique.
Normal TAV Hemodynamics Categorized on the Basis of Native Annular Dimensions
TAVR has reminded us of the limitations of standard 2-dimensional measurements of an LVOT that is often not circular. Thus, 3-dimensional measured perimeter, area, or diameter is superior for annular sizing. Computed tomography (CT) is now the default approach for pre-procedural annular measurement with 3-dimensional echo-based sizing a backup when CT is impossible or suboptimal. In addition, the concept that the ultimate TAV EOA depends on not only the nominal valve size, but also deployment characteristics; the restraining effects of the annulus and/or calcium has emerged as a unique consideration for TAVR. It has even been suggested that the stent may remodel over time. For rigid surgical prostheses, valve size alone determines EOA. Based on these concepts, the authors have proposed an alternative to valve type and size to define normal hemodynamic parameters based on CT-measured native annular diameter. This is a useful concept and it would have been interesting if the authors had commented on the degree to which the normal values based on valve size are consistent with those based on pre-implantation native annular diameter. If there were discordance, it would potentially leave the user with values that are normal by 1 approach but abnormal by the other. It is unclear why the study did not use CT perimeter or area rather than a single diameter for this analysis.
Comparison of Normal Values for Different Valve Iterations
It is unfortunate that the authors did not provide between-iteration comparisons. They explain this on the basis of differences in patient cohorts and core laboratories. It is unlikely, however, that cohorts that differed largely on the basis of surgical risk would confound comparisons and surely the argument for the superiority of core laboratory measurements is based on the presumed consistency of the measurement approaches that should have been standardized among the core laboratories. As different valves are introduced such comparisons would be of interest in future studies.
There are other limitations that deserve attention. First is the small number of self-expanding valves, in particular the 23-mm CoreValve and Evolut R valves. The variation in approach to measurement (in-stent vs. pre-stent in some) may also be an issue. Second, as with native valves, prosthetic valve function must be assessed in the context of loading conditions and stroke volume. Information is not provided as to blood pressure or ejection fraction, and it is reasonable to assume that measurements included those from patients with hypertension or systolic dysfunction. This detail would be informative and should be included in future reports of TAVs. More important, aortic regurgitation may influence gradients and it would be reasonable to have excluded any subject with more than mild regurgitation because the paper purports to provide data on normally functioning valves. Although this most likely was the case, it is not explicitly stated. Finally, it would have been reasonable to compare the results here with prior reports. Although a detailed comparison is beyond the scope of this editorial, a quick review suggests that the results here are largely concordant with earlier reports.
This paper provides important tools for the echo-Doppler assessment of transcatheter aortic valve function. By highlighting flow characteristics that differentiate transcatheter from surgical valves, the authors provide a well-illustrated guide for optimized standardized approaches to calculating LVOT stroke volume and EOA. Normal values based on prosthetic valve type and size as well as baseline annular diameter complement prior reports, the majority of which did not apply the rigor associated with core laboratory adjudication.
↵∗ Editorials published in JACC: Cardiovascular Imaging reflect the views of the authors and do not necessarily represent the views of iJACC or the American College of Cardiology.
Dr. Gillam directs an imaging core laboratory with research contracts with Edwards Lifesciences and Medtronic, but she does not receive direct compensation for this work. Dr. Marcoff participates in the work of this core laboratory, but he does not receive direct compensation for this work.
- 2019 American College of Cardiology Foundation
- Hahn R.T.,
- Leipsic J.,
- Douglas P.S.,
- et al.
- Winter M.P.,
- Zbiral M.,
- Kietaibl A.,
- et al.
- Gillam L.D.,
- Leipsic J.,
- Weissman N.J.
- Corresponding Author
- Importance of Core Laboratory-Adjudicated Results and a Standardized Approach to Image Acquisition
- Relation Between Valve Size and Hemodynamics
- Normal TAV Hemodynamics Categorized on the Basis of Native Annular Dimensions
- Comparison of Normal Values for Different Valve Iterations
- Study Limitations