Statistics from Altmetric.com
Calcific aortic stenosis (AS) is characterised by an initial inflammatory phase, with endothelial damage, lipoprotein infiltration and oxidative stress. This stage is followed by a disorganisation of the extracellular matrix with an overproduction of collagen fibres and a mineralisation phase in which valvular interstitial cells differentiate into an osteoblast-like cell type, leading to microcalcification analogue to skeletal bone formation. This self-perpetuating cycle of progressive fibrocalcific remodelling leads to macrocalcification, increases valve rigidity and therefore narrows the effective orifice area, increasing left ventricular afterload and ultimately provoking myocardial damage and symptoms (figure 1).1 Valvular fibrosis, although understudied, appears to play a major role in younger, bicuspid valve and female patients and is one of the reasons sex-specific thresholds are used to define severe AS by CT aortic valve calcification (CT-AVC) quantification (2000 Agatston units (AU) for men and 1200 AU for women).2 The pathogenic model depicted in figure 1 is not necessarily followed in a serial fashion, and although the advance of the disease is inexorable, the rate of AS progression remains largely unpredictable. However, some markers of rapid progression have been described, such as systolic hypertension,3 bicuspid morphology4 and metabolic syndrome.5
The road to development of pharmacological treatments to either prevent or slow down AS progression has been uneven, but it is still a critical goal that may obviate aortic valve replacement altogether, which remains to date the only available treatment. An ideal non-invasive imaging biomarker for clinical trials of potential AS disease-modifying treatments should (1) allow tracking of subtle changes in disease progression over time; (2) allow accurate diagnosis of AS severity; (3) have prognostic impact on clinical outcomes; and (4) have excellent intra-rater, inter-rater and test–retest reproducibility.
Doppler echocardiography is currently the most universally used method to assess the severity and progression of AS, as well as its consequences on the myocardium. However, up to 40% of patients with AS present with discordant findings on echocardiography (ie, a low mean gradient (<40 mm Hg) or peak aortic velocity (<4 m/s) and a small aortic valve area (<1 cm2)), which raise uncertainty about the true severity of AS.6 The use of CT-AVC quantification has emerged as a robust, flow-independent, semiquantitative method for evaluation of AS anatomical severity. Furthermore, it has been incorporated in the European Society of Cardiology (ESC)/European Association for Cardio-Thoracic Surgery (EACTS) guidelines for AS true severity adjudication in low-gradient AS.7 However, its potential for assessing disease progression as compared with echocardiography has never been formally tested.
In this issue of Heart, Doris et al 8 present an insightful study in which they compared the ability of CT-AVC to track AS progression with echocardiography. The hypothesis was that CT-AVC would provide a more sensitive measure of disease progression than echocardiography and thus would be useful both for clinical follow-up of patients with rapidly progressing AS and as an imaging endpoint in clinical trials of novel therapies designed to slow disease progression. Overall, scan–rescan reproducibility was markedly superior for CT-AVC as compared with echocardiography. Interestingly, the modest annualised changes observed by echocardiography were well within the measurement error, as evidenced by the Cohen’s d statistic, which was 0.71 for peak velocity, 0.66 for mean gradient and 0.59 for aortic valve area. The best performing echocardiographic method (although still within measurement error) was dimensionless index, with a Cohen’s d statistic of 1.41. On the other hand, annualised changes for CT-AVC were significantly higher than the measurement variability (Cohen’s d statistic=3.12). Unfortunately, the authors did not provide the proportion of women on each subgroup and no adjustment was made to account for sex differences in CT-AVC, which is known to influence CT-AVC progression.9 As expected from the Cohen’s d statistics, the group size required both to detect annualised disease progression and potential treatment effect of new therapies was more than 10-fold smaller for CT-AVC than for echocardiographic measurements.
One of the main strengths of the study is the assessment of scan–rescan reproducibility (as opposed to independent reading of preacquired images) for both echocardiography and CT-AVC. The article provides an interesting proof of the intrinsic variability of echocardiographic assessment of AS, as the risk of significant haemodynamic variations between scans was practically reduced to zero by performing both echocardiographic studies on the same day. In real life, however, changes in haemodynamic status (especially due to variable loading conditions) can have a major impact on the estimation of peak velocity, mean gradient, and to a lower extent aortic valve area (AVA). Regarding some limitations of the study, the fact that AVA reproducibility was so poor underscores the need for a meticulous approach to AVA calculation in AS. Special attention must be made to the measurement of the left ventricular outflow diameter (LVOT), which can have a major impact on stroke volume estimation and AVA calculation. Measuring LVOT diameter at the annulus (and not 5–10 mm below as suggested by guidelines) has proven to be more reproducible and provide better agreement with invasive measurements and cardiovascular magnetic resonance.10
The rate of exclusion in CT-AVC was rather high (16%), which might have biased the results to favour CT-AVC. Although it would have been interesting to have done the repeated echocardiograms and CT-AVC in the same patients, AS severity was comparable and is unlikely to have biased the results significantly. Finally, an analysis that takes into account sex differences would have been insightful, especially for CT-AVC.
In summary, CT-AVC is clearly a more robust imaging biomarker for AS progression than echocardiography, as it accurately diagnoses severe AS,2 provides independent prognostic value,11 tracks subtle changes in disease progression and has an excellent test–retest reproducibility, as demonstrated in this study. Thus, we believe the evidence is compelling to consider CT-AVC one of the most attractive imaging biomarkers for cohort studies evaluating AS progression or randomised controlled studies assessing the impact of potential novel disease-modifying therapies.
In clinical practice, however, despite its higher variability (multi-window interrogation, acoustic window, changes in loading conditions, arrhythmia and so on) and therefore lower ability to detect subtle changes in AS progression, Doppler echocardiography remains an extremely versatile imaging modality. It not only allows estimation of the actual haemodynamic impact of AS, but also allows comprehensive assessment of the consequences on the rest of the heart (left ventricular systolic and diastolic function, assessment of the right heart, concomitant valvular disease and so on), whereas CT-AVC only provides information on the quantity of cardiac calcification. Thus, CT-AVC has definitely a complementary role in defining AS severity in discordant-grading patients (figure 2), but cannot be used alone. As proposed by the present study, serial follow-up with CT-AVC in selected patients (poor echogenicity, systolic hypertension, low flow, suspected rapid progressor and so on) may improve a clinical surveillance (ie, ‘watchful waiting’) strategy (figure 2). However, before adopting a ‘CT to all’ strategy that would increase costs and complexity, such strategy should demonstrate to provide significant benefits on clinical outcomes, and a detailed analysis of cost-effectiveness of both surveillance strategies should be performed.
The authors would like to thank Véronic Tremblay for her help in figure preparation.
Contributors EG drafted the first version of the editorial. M-AC revised the draft of the editorial.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Patient consent for publication Not required.
Provenance and peer review Commissioned; internally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.