What Our Eyes See Is Not Necessarily What Our Heart Feels

The accurate assessment of hemodynamic significance is crucial for clinical decision making in patients with aortic stenosis (AS) [1]. Over the past decades, echocardiography has become the clinical standard for the evaluation of AS severity and the most frequently used index for this purpose is the aortic valve area (AVA). According to ACC/AHA recommendations [2], the stenosis is considered mild if the AVA is >1.5 cm2, moderate if ≤1.5 cm2 and >1.0 cm2, and severe if ≤1.0 cm2. The AVA can be measured by planimetry on the bidimensional images obtained by transthoracic (TTE) or transesophageal (TEE) echocardiography. It can also be measured by Doppler echocardiography (Doppler TTE) with the use of the continuity equation. Unfortunately, the accuracy of the echocardiographic measurement of AVA may be affected by several factors. For example, inadequate acoustic windows hinder accurate measurements in up to 10% of subjects. Furthermore, bright echoes and shadowing caused by the valvular calcifications present in many patients with severe AS may render difficult or unreliable the planimetric measurement of the valve orifice area. When using the continuity equation method, the AVA may be overestimated if the Doppler beam is not well aligned with the direction of the transvalvular flow jet. Moreover, patients with severe AS generally have a marked concentric left ventricular (LV) hypertrophy that is often associated with a protrusion of the interventricular septum into the LV outflow tract. This may cause a subvalvular acceleration of the flow and thus an overestimation of the AVA. On the other hand, an underestimation of the left ventricular outflow tract diameter may lead to an underestimation of the AVA. Alternatively, the AVA can be measured by catheterization with the use of the Gorlin formula. Unfortunately, this method is invasive and is associated with a substantial risk of cerebral embolism in the presence of severe AS [3, 4]. Moreover, it is also subject to significant measurement errors that may be related to nonsimultaneous recording of LV and aortic pressures (pullback method), to inaccurate measurement of transvalvular flow, and to the occurrence of reverse pressure gradient at the end of systolic ejection [1, 4, 5]. There is clearly much room for improvement, creating a strong impetus for the development of new noninvasive methods that could provide an accurate estimate of the AVA in AS patients. Such a method would be particularly useful in the situations where the echocardiographic measurement of the AVA is not feasible or is potentially inaccurate or when there are inconsistencies between the measured AVA and the symptomatic status of the patient. Cardiovascular magnetic resonance imaging (CMR) is particularly appealing because of its ability to noninvasively measure precisely and reproducibly cardiac anatomy and physiology. Cine sequences such as the steady-state free-precession method provide superior temporal and spatial resolution, with less susceptibility to blood flow artifacts and shorter scan times compared to previous sequences. Furthermore, CMR is appealing compared to Doppler echocardiography because of its ability to image the valve and measure blood flow velocities and derive pressure gradients in precise scan planes, rendering the technique less susceptible to ‘interrogation angle error’. Additionally, imaging is not impaired by ‘poor window quality’ and much less susceptible to artifacts associated with calcified AS. Like echocardiography, CMR offers a complete assessment including cardiac function and chamber dimensions, with the added ability of accurately reporting ventricular mass. It must be noted, however, that in its current iteration CMR has not yet reached the rapid real-time processing abilities of Doppler ultrasound for which TTE remains the foremost screening tool available for initial assessment of valvular disease. In this issue of Cardiology, Malyar et al. [6 ]compared the AVA measured by CMR with that measured by TTE or TEE. The results show a strong correlation and agreement between CMR and TEE for the measurement of AVA but a weaker correlation with Doppler TTE. These findings complement previous work demonstrating a strong correlation between CMR and TEE for the measurement of AVA by planimetry [7], but a weaker correlation between CMR measurement of AVA by planimetry and Doppler TTE calculation of AVA by the continuity equation [8].The AVA may be measured by several methods, including (1) two-dimensional ultrasound planimetry; (2) Doppler TTE derived by the continuity equation; (3) catheter derived by the Gorlin formula; (4) CMR planimetry, and (5) velocity-encoded CMR derived by the continuity equation. This latter method was not used by the authors in the present study. One important question is then: do these methods really measure the same thing? The planimetry methods (No. 1 and 4) determine the anatomic orifice area (AOA) that is the geometric area of the valve orifice (fig. 1). This is the opening that our eyes see on the TTE, TEE or CMR images. On the other hand, the derivational methods using the Gorlin or continuity equations (No. 2, 3 and 5) calculate the effective orifice area (EOA) of the valve, which represents the cross-sectional area of the transvalvular flow jet at the level where the flow contraction is maximal (fig. 1). Indeed, when the blood flow passes through a stenotic orifice, there is a contraction and acceleration of the flow and the area where the flow jet is the smallest is called the vena contracta. The TTE, TEE or CMR planimetric methods provide a measurement of the anatomic area of the stenotic orifice, i.e. of the AOA, whereas Doppler TTE gives an estimate of the cross-sectional area of the vena contracta, i.e. of the EOA. Accordingly, velocity-encoded CMR measurements correlate strongly with those obtained by Doppler TTE [9, 10].Hence, the first important message is that the different methods used to estimate the AVA do not measure the same parameter [11, 12]. In particular CMR and TEE measure the AOA, whereas Doppler TTE and velocity-encoded CMR measure the EOA (fig. 1). The next question is then: is there a good correlation and agreement between these 2 parameters? The AOA and the EOA are related by the following equation:EOA = AOA × CCwhere CC is the contraction coefficient. The contraction coefficient essentially depends on the relationship between the size of the LV outflow tract and the size of the valve orifice (i.e. the AOA) and also importantly on the shape of the valve inflow [11,12,13,14]. The CC is close to 0.6–0.7 when the valve inflow plane is flat and perpendicular to flow direction as may occur in severe calcified AS. In this situation, the EOA is much smaller than the AOA (table 1). On the other hand, the CC is close to 1.0 in funnel-type valve inflow shape as may occur in normal aortic valves and in this situation, the EOA is close to AOA [11,12,13,14]. Hence, the AOA overestimates the EOA to a different extent depending on the valve inflow shape. The understanding of these discrepancies between AOA and EOA is useful to explain the results obtained in the present study [6]. There was indeed an excellent correlation and agreement between the AVA measured by CMR and that measured by TEE because these 2 methods measure the same parameter, i.e. the AOA of the valve, whereas the correlation was lower between CMR-derived AVA and Doppler TTE-derived AVA. This is not surprising given that these 2 methods do not measure the same parameter. The observation that CMR-derived AVA overestimated Doppler TTE-derived AVA is consistent with the concept that for a given AOA (i.e. AVA measured by CMR) of 1.0 cm2, the EOA (i.e. AVA measured by Doppler TTE) may vary from 0.6 to 1.0 cm2 depending on the size of the LV outflow tract and valve inflow geometry [12,13,14]. Consistently, a similar discrepancy was observed between TEE AVA (AOA) and Doppler TTE AVA (EOA). Another factor that may contribute to the discrepancy between AVA measured by CMR or TEE and AVA measured by Doppler TTE is the timing of the measurements. CMR and TEE are similar in measuring the AOA at a single time frame during the cardiac cycle, i.e. at peak systole, whereas TTE measures the average EOA over the whole ejection period. To this effect, it has been demonstrated that the AOA and EOA may vary markedly during ejection, especially in patients with severe AS, which implies that maximum and average values of EOA or AOA are different [15, 16]. This factor may thus also have contributed to the fact that CMR- or TEE-derived AVA generally overestimates the Doppler TTE-derived AVA given that the 2 former methods reflect the maximum AVA as opposed to the TTE method that reflects the average AVA. The findings presented above support the notion that the AOA and the EOA are different parameters and that differences up to 60% may be observed depending on the valve inflow geometry and flow rate. The next important question is then: what is the most relevant parameter to evaluate AS severity and predict clinical outcome? In this regard, it must be remembered that the transvalvular pressure gradient and thus the LV workload are essentially related to the EOA of the valve and not to its AOA [11, 12, 17]. As illustrated in table 1, the EOA is superior to the AOA to assess the increased burden imposed by the stenosis on the LV. From a clinical standpoint, the EOA can be expected to provide greater insight compared to the AOA for the assessment of stenosis severity and the prediction of clinical outcomes. Hence, a logical extension of the work presented by Malyar et al. [6 ]in this issue ofCardiology would be to refine and validate CMR-based methods to measure the EOA of the valve, i.e. the cross-sectional area of the flow jet at the level of the vena contracta [10]. This would allow the CMR specialist to measure the ‘physiological’ AVA that is the main determinant of the hemodynamic burden imposed by the valve on the LV.In a substantial proportion of the patients included in the present study, the classification of the stenosis severity was different depending on the method used to measure the AVA [6]. In this regard, it should be emphasized that the ACC/AHA guidelines for defining AS severity were initially established mainly based on data obtained from catheter measurements as well as on clinical outcomes in relation to these measurements [2]. The same values for AS severity (e.g. <1.0 cm2) were then extended to echocardiographic data and now to CMR data on the assumption that catheter-, TEE-, Doppler TTE-, CMR-, and velocity-encoded-CMR-derived AVAs were equivalent parameters and indeed, the guidelines make no distinction between catheter, echocardiographic, or CMR measurements. The same threshold values are applied to grade stenosis severity regardless of the method used to measure the AVA. As reported in the present study as well as in previous studies [1, 11, 12, 17], these parameters are not necessarily equivalent and discrepancies up to 50–60% may be observed. Hence, the present guidelines, based mostly on AVAs measured during catheterization, may not be directly applicable to measurements made by echocardiography or CMR and may result in misclassification of severity, thus affecting clinical management.In conclusion, what our eyes see on the CMR images described in the current study is the AOA but what the heart of our patient feels is the hemodynamic burden that results from the EOA. The next logical step should thus be to refine CMR-based methods to accurately assess the EOA. Dr. Philippe Pibarot holds the Canada Research Chair in Valvular Heart Diseases, Canadian Institutes of Health Research, Ottawa, Ont., Canada.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.