Aortic valve replacement is indicated in patients with severe aortic stenosis (AS) with symptoms or depressed left ventricular (LV) ejection fraction. Doppler echocardiography is the primary imaging modality used to assess the severity of AS, and this modality therefore plays a crucial role in the therapeutic management of patients with AS. In a well-reasoned and elegantly written viewpoint, Shah rightfully states that the quantitation of AS severity by Doppler echocardiography is not a settled science, and he underlines some very important limitations of this technique, which merit further discussion.
Which Is the Best Doppler Echocardiographic Parameter to Grade AS Severity?
The mean transvalvular gradient and aortic valve area (AVA) are the two main echocardiographic parameters that are used to grade AS severity. A stenosis is considered severe when the mean transvalvular gradient is >40 mm Hg (or the peak aortic jet velocity is >4 m/sec), whereas a stenosis is considered nonsevere when the mean gradient is <40 mm Hg and the AVA is >1.0 cm 2 ( Figure 1 ). However, when the gradient is >40 mm Hg but the AVA is <1.0 cm 2 , the stenosis severity is uncertain, and additional tests are generally required to confirm severity. This situation occurs in up to 30% to 40% of patients with AS and is often referred as to “discordant grading” or “low-gradient AS” in the literature.
The main disadvantage of the gradient is that it is highly flow dependent, and this phenomenon explains why the discordant grading pattern (small AVA but with low gradient) often occurs in patients with low flow. Indeed in the presence of low flow (generally defined as a stroke volume index < 35 mL/m 2 ), the gradient may be “pseudonormalized” and may thus underestimate stenosis severity. On the other hand, AVA, although less flow dependent than the gradient, may also be smaller because of the low transvalvular flow rate, thus resulting in overestimation of stenosis severity (i.e., pseudosevere stenosis). AVA calculated by the continuity equation method is the “effective” orifice area (i.e., the cross-sectional area of the transvalvular flow jet), whereas AVA measured by planimetry represents the “anatomic” orifice area. Because of the flow contraction that occurs downstream of the valve orifice, the effective AVA is generally smaller than the anatomic AVA. In the presence of low flow, the valve leaflet may not be completely open, and therefore both the anatomic and effective AVAs may be pseudosevere ( Figure 1 ).
Hence, a Doppler echocardiographic examination performed at rest often does not provide an accurate estimation of AS severity in the presence of low transvalvular flow. In such a situation, low-dose (up to 20 μg/kg/min) dobutamine hemodynamic stress echocardiography is generally recommended to differentiate true from pseudosevere AS and to guide therapeutic management ( Figure 1 ).
Shah also underlined that besides AVA, the mean transvalvular flow rate (not only stroke volume) is the main hemodynamic determinant of the gradient. The mean flow rate is directly related to stroke volume and inversely related to LV ejection time. Hence, the gradient is dependent not only on LV outflow (i.e., LV outflow tract [LVOT] stroke volume) but also on chronotropy. For a given AVA and stroke volume, the gradient is smaller when heart rate is slower (i.e., LV ejection time is longer). Furthermore, the gradient is also dependent on arterial hemodynamics. Several studies indeed reported that the gradient may be blunted in the presence of reduced arterial compliance and systemic arterial hypertension.
The AVA has the advantage of being less dependent on stroke volume, heart rate, and arterial hemodynamics than the gradient. On the other hand, as emphasized by Shah, AVA is more prone to measurement errors ( Figure 1 ). AVA is indeed calculated with the use of the continuity equation, which includes three measured parameters: the LVOT velocity-time integral, the LVOT diameter, and the aortic flow velocity-time integral. Among these parameters, the most “vulnerable” is certainly LVOT diameter because, first, it is technically challenging to measure, especially in elderly patients with calcific AS, and second, it is squared in the continuity equation. Furthermore, this equation assumes that the LVOT cross-section is circular, whereas several studies using three-dimensional (3D) echocardiography, multidetector computed tomography (MDCT), or cardiac magnetic resonance (CMR) revealed that the LVOT shape is in fact elliptical in most patients. Two-dimensional (2D) echocardiography measures the anteroposterior diameter of the LVOT, which is generally the smaller diameter of the ellipse. Hence, the calculation of LVOT area from LVOT diameter measured by echocardiography may lead to an underestimation of the actual stroke volume, which in turn would result in a calculated AVA that is smaller than the actual AVA.
Besides AVA and gradient, other echocardiographic parameters may be used to grade AS severity. The peak aortic jet velocity has the advantage of being a simple and reproducible measure of AS severity. However, it faces the same limitations as the gradient in terms of flow dependency. The Doppler velocity index or dimensionless ratio, which is calculated by dividing the LVOT velocity-time integral by the aortic flow velocity-time integral, can be used as an adjunctive parameter to assess AS severity. In a patient with discordant grading, the finding of a Doppler velocity index >0.25 should raise the suspicion of an error in the measurement of the LVOT diameter with ensuing underestimation of AVA ( Figure 1 ). However, this index, like AVA, may be pseudosevere in the presence of low flow. Furthermore, it is also dependent on LVOT size and may underestimate stenosis severity in patients with small LVOTs and overestimate severity in those with large LVOTs.
In summary, there is unfortunately no perfect echocardiographic parameter to quantitate AS severity; relying on one single parameter may be misleading and may result in an erroneous grading of AS severity.
Which Is the Best Modality to Confirm Stenosis Severity in Patients with Discordant Grading?
Hybrid Methods to Measure AVA
To overcome the aforementioned pitfall of 2D echocardiography for the measurement of LVOT area included in the continuity equation, some investigators have proposed using a hybrid approach in which LVOT area is measured by MDCT, CMR, or 3D echocardiography and the LVOT and aortic velocities are determined using Doppler echocardiography ( Figure 1 ). However, these hybrid methods are certainly not a panacea, and in fact the few studies that attempted to validate these methods found that compared with AVA measured by standard echocardiography, hybrid AVA failed to improve the correlation with the gradient or the prediction of clinical outcomes. These findings may be explained by the fact that the hybrid methods are also subject to a number of pitfalls and limitations:
Whereas 2D echocardiography tends to underestimate LVOT area, the hybrid methods, on the other hand, may result in an overestimation of the area. In an in vitro study using phantoms, Tsang et al. reported that MDCT overestimates LVOT area compared with CMR. Furthermore, in vivo, the use of an imaging plane that is not perpendicular to the LVOT may result in an overestimation of the LVOT cross-sectional area, regardless of the imaging modality being used. This limitation of the hybrid methods was well illustrated in a study by Kamperidis et al. , in which in a series of patients who were confirmed to have symptomatic severe AS by a multidisciplinary heart team and who underwent transcatheter aortic valve replacement, 16% would have been reclassified with nonsevere AS on the basis of hybrid (MDCT-echocardiography) AVA. Among these patients reclassified as nonsevere by hybrid AVA, several had high gradients. In our experience, it is not rare to see patients with high gradients and undoubtedly severe AS but nonetheless hybrid AVAs > 1.0 cm 2 or even > 1.5 cm 2 . In a symptomatic patient, underestimating AS severity may result in underuse and/or delay of aortic valve replacement, which may in turn negatively affect the outcome. Hence, the hybrid methods recently proposed to measure AVA cannot be considered as a gold standard for the quantitation of AS severity.
The cut-point values of AVA proposed in the guidelines to grade AS severity were originally established with AVA measured by Doppler echocardiography and have been extensively validated with outcome data. Given that the hybrid methods measure larger AVAs compared with those obtained by standard echocardiography, the cutoff value of 1.0 cm 2 (or 0.6 cm 2 /m 2 ) used to define severe AS cannot be applied to the hybrid methods. Outcome studies are needed to validate the severity cut points of AVA that should be used for hybrid methods. To this effect, Clavel et al. demonstrated that for AVA obtained by echocardiography, the best cut-point value to predict mortality in patients with AS managed conservatively is close to 1.0 cm 2 . However, the optimal cut-point value was larger (1.2 cm 2 ) for AVA measured by the hybrid (MDCT-echocardiography) method. Hence, larger severity cut-point values may have to be used for AVA measured by the hybrid MDCT-echocardiography method.
Before considering a hybrid method to measure AVA, it would make sense to first attempt to optimize the measurement of LVOT area by 2D echocardiography. On the basis of the physiologic rationale that LVOT diameter should be measured at approximately the same location used to measure the flow velocity, the 2009 American Society of Echocardiography guidelines for the quantitation of AS severity by Doppler echocardiography suggested to measure LVOT diameter 5 to 10 mm below the aortic annulus. However, the cross-sectional shape of the LVOT at this level is more elliptical, more irregular (because of the frequent septal bulge), and more dynamic (greater changes between diastole and systole) than at the level of the aortic annulus. From a historical perspective, it should be emphasized that the continuity equation method to measure AVA by echocardiography has been initially described with LVOT diameter measured at the level of the aortic annulus. And some recent studies using AVA measured by left heart catheterization actually confirmed that this original method using LVOT diameter at the aortic annulus is actually more accurate and reproducible than the method using LVOT diameter measured 5 to 10 mm below the annulus. To measure LVOT diameter at the level of the annulus, magnified views providing the largest annular dimension should be used. In a tricuspid aortic valve, the maximum diameter of the annulus is the plane that bisects the right coronary cusp, anteriorly, and the commissure between the left coronary and noncoronary cusps, posteriorly. Hence, if two cusps are well visualized in the parasternal long-axis view, this may not yield the largest and true LVOT diameter.
The hybrid methods are more cumbersome and more expensive than echocardiography because they require the performance of two different imaging examinations to measure a single stenotic parameter (i.e., AVA). A simpler and more cost-effective approach to determine hybrid AVA would be to measure LVOT area by 3D echocardiography during the standard Doppler echocardiographic examination. However, the suboptimal image quality often achieved with 3D transthoracic echocardiography may limit the accuracy of the LVOT area measurement, and transesophageal echocardiography may be required to obtain an accurate assessment of the hybrid AVA. Hence, although the hybrid methods may be useful to corroborate the stenosis severity and eventually may help in therapeutic decision making in symptomatic patients with discordant grading on echocardiography ( Figure 1 ), these methods are likely not relevant and not applicable for the routine assessment and follow-up of AS severity.
Low-Dose Dobutamine Hemodynamic Stress Echocardiography
In the setting of low-flow states, neither AVA nor the gradient measured at rest by any modality (echocardiography, catheterization, MDCT, CMR, or a hybrid method) is accurate to quantitate stenosis severity and differentiate true from pseudosevere stenosis. Low-dose dobutamine stress echocardiography can be used to increase flow to remeasure AVA and gradient, ideally at a normal flow rate. If the mean transvalvular gradient increases to >40 mm Hg, the stenosis is then considered truly severe, whereas if the mean gradient remains below 40 mm Hg and the AVA increases above 1.0 cm 2 as flow increases, then the stenosis is considered pseudosevere. However, a large proportion of patients have only a modest increase in flow during dobutamine infusion, and they thus remain below the normal range of flow rate. As a consequence, the discordant grading observed at rest (i.e., small AVA with low gradient) often persists at dobutamine stress. To address this challenging situation, the investigators of the Truly or Pseudo Severe Aortic Stenosis (TOPAS) study proposed using projected AVA at normal flow rate. This parameter is in fact a projection of what would be the AVA at normal flow rate by calculating the slope of AVA-flow relationship measured during dobutamine stress. This new parameter has been validated in the context of both classical (with low LV ejection fraction) and paradoxical (with preserved ejection fraction) low-flow, low-gradient severe AS by comparison with in vitro, surgical, pathologic, and clinical outcome data. Shah is right in stating that because of its complexity, this new parameter has not been yet widely used in clinical practice. The TOPAS investigators proposed a simplified version of the projected AVA that can be obtained by a simple calculation using the AVA and flow rate measured at rest and at peak dobutamine stress. Projected AVA should be calculated only when the AVA-gradient discordance persists at the end of dobutamine stress echocardiography. Hopefully, the availability of the simplified projected AVA will facilitate wider implementation of this parameter in practice.
Shah proposes using the flow-gradient relationship assessed during dobutamine stress to confirm stenosis severity in patients with low-gradient AS. This is certainly an interesting suggestion that deserves further study. However, there are several pitfalls that can already be anticipated with this approach. First, as for the projected AVA, this approach would require the measurement of LVOT velocity, aortic valve velocity, and LV ejection time at rest and during dobutamine stress, which would necessarily imply larger measurement variability. Second, given that a large proportion of patients have limited flow reserve and thus do not normalize their flow with dobutamine infusion, it would be necessary to use an approach similar to the projected AVA and calculate “the gradient projected at normal flow rate.” However, the estimation of this projected parameter may be complex given that, as opposed the flow-AVA relationship, the flow-gradient relationship is not linear.
Other Alternative Methods to Grade AS Severity
Valve leaflet calcification is the main culprit lesion of AS, and therefore the quantitation of aortic valve calcification has been proposed to corroborate AS severity. MDCT is currently the most widely used method to measure aortic valve calcification. However, new echocardiographic methods are currently being developed to quantitate aortic valve calcium load. Valve calcification measured by MDCT has been shown to well predict AS hemodynamic severity, AS progression rate, and clinical outcomes. However, this method is not a panacea either. First, it measures the anatomic rather than the hemodynamic severity of the disease. Second, it does not measure valve leaflet fibrosis, which, besides calcification, also contributes to the valve stenosis. To this effect, it has recently been reported that MDCT may underestimate the hemodynamic severity of AS, especially in younger patients with bicuspid valves.
In summary, there is no gold-standard modality for quantitating AS severity. Each modality has strengths and limitations that need to be recognized and integrated in the interpretation of the various parameters measured with these modalities.