A Clinician’s Guide to the Changing Aortic Stenosis Landscape: Updates in Aortic Stenosis Diagnosis, Surveillance and Management

This review aims to update healthcare providers on contemporary diagnostic and treatment information pertaining to aortic stenosis. The prevalence of aortic stenosis continues to increase, and so does the burden of treatable disease. This has important implications for healthcare systems and the economy. Accurate classification of aortic stenosis severity and determining optimal treatment timing remains a prime challenge to treating physicians. Furthermore, the drastic uptake of transcatheter aortic valve replacement has generated novel issues pertaining to younger patients whose post-intervention survival now exceeds device durability. The keys to optimizing patient outcomes are delivering accuracy in disease classification, treatment selection and timing and procedural planning.


Introduction
Calcific aortic stenosis (AS) is a major contributor to the global burden of valve disease and is increasing in prevalence [1].The natural history of AS is characterized by a long latent asymptomatic period, followed by an abrupt and rapid increase in mortality once symptoms develop, usually in the seventh to ninth decades of life [2].Without treatment, the onset of symptoms heralds a life expectancy thenceforth of about 3 years.Even with medical advances and the introduction of minimally invasive approaches to replace the diseased aortic valve, determining exactly when to do so has continued to be the key clinical dilemma facing clinicians for many decades.Up until recent times, those with mild, moderate, or asymptomatic disease were thought to have a "good outlook" [3], but deeper pathophysiological understanding coupled with large contemporary population studies have highlighted that this is not the case [4••].With improved imaging techniques, poor survival and prospects of intervention can now be predicted well before the development of symptoms [5].Herein, we review contemporary data pertaining to AS, with a focus on diagnostic criteria, patient surveillance, and treatment options.The aortic landscape is rapidly evolving, and in this review, we provide a clinician's guide to navigating this complex space.

Disease burden
The global prevalence of AS continues to steadily increase [1,6].The agestandardized prevalence of calcific aortic valve disease was 116.3 cases per 100,000 people in 2019, more than doubling since 1990 [7].AS is particularly prevalent in high-income countries, partly driven by its strong correlation with age; pooled meta-data suggest contemporary prevalence exceeds 10% in the population aged over 75 years [8].This creates significant challenges for healthcare infrastructure and resource allocation, given the growing treatable burden of disease [9].With advancements in economies in developing nations, the prevalence of age-related AS will increase [6].Although traditional atherosclerotic risk factors, such as smoking and hyperlipidemia [10], have been shown to correlate with the incidence of AS, their modification with pharmacotherapy has not been successful so far in preventing disease progression.Replacement of the aortic valve, either surgically or via the transcatheter approach, is the only treatment which improves survival.

Symptoms
Progressive stenosis of the aortic valve initiates cellular changes in the left ventricular myocardium due to pressure over-loading.The chronic pressure load leads to reactive myocardial hypertrophy, which helps to maintain cardiac output.As the disease progresses, symptoms manifest due to a combination of hypertrophy-related subendocardial ischemia, fibrotic remodelling of the ventricle, and ultimately insufficient cardiac output to meet peripheral demand while preserving normal left ventricular filling pressures.Accordingly, the cardinal symptoms of significant AS -exertional chest pain, dyspnea, or syncope -herald an advanced disease stage with poor survival in those untreated.However, the long-held belief that AS follows a benign trajectory prior to the onset of symptoms or overt ventricular dysfunction has been questioned recently [4••, 11, 12].Symptom onset is insidious and unpredictable.Early in the disease course, symptoms often require provocation with objective exercise testing [13•].Although the risk of sudden death during the asymptomatic phase is theoretically low provided symptoms are promptly reported when they arise, "watchful waiting" in asymptomatic severe AS is associated with worse outcome than early surgical intervention [14,15].

Diagnostic criteria and inherent challenges
Echocardiography forms the cornerstone of contemporary diagnostic assessment.Current American College of Cardiology (ACC)/American Heart Association (AHA) and European Society of Cardiology (ESC)/European Association of Cardiothoracic Surgery (EACTS) guidelines classify severe AS as a mean transaortic gradient ≥ 40 mmHg (calculated with the modified Bernoulli equation), a peak transaortic velocity ≥ 4 m/s, or an aortic valve area (AVA; calculated by the continuity equation) ≤ 1.0cm 2 [13•, 16•].Up to 50% of patients with severe AS will demonstrate discordance in one or more of these echocardiographic metrics [17][18][19], due to a combination of mathematical assumptions and physiological factors such as ventricular geometry, contractility, and transaortic flow state.Additionally, these major hemodynamic metrics do not experimentally yield an inherently consistent severity classification.Minners et al. eloquently demonstrated that a mean transaortic pressure gradient of 40 mmHg yields a Gorlin-predicted AVA of 0.81 cm 2 and continuity-calculated AVA of 0.75cm 2 [20•].Conversely, an AVA of 1.0cm 2 corresponds to a Gorlin-predicted mean pressure gradient of 26 mmHg and echocardiographically measured mean pressure gradient of 22.8 mmHg.As such, a significant number of patients are discordantly classified as severe by the AVA criterion, but not by the pressure criterion.This phenomenon is the most common discordant pattern [21], referred to as "low gradient" severe AS [18].Discordance complicates treatment decision-making, particularly in patients without symptoms or those with other comorbidities to which symptoms might be ascribed.Efforts to clarify the prognostic implications of the discordance patterns has led to the sub-classification of severe AS by gradient (high or low) and by flow status (normal or low).
Discordance is particularly prominent in patients with a low cardiac output state, characterized by low stroke volume [20•].Reduced stroke volume is commonly referred to as a "low flow" state, defined as an indexed stroke volume (SVi) < 35mls/m 2 [13•, 18].However, this is somewhat of a misuse of the term "flow" which actually represents change in volume per unit time.Cardiac output (CO) is a measure of flow but does not define flow status [13•].A related, though distinct flow metric is transvalvular flow rate (Q), which quantifies the stroke volume with direct relation to the period of systolic ejection, rather than the entire cardiac cycle [17].Flow state, on which the calculation of valve metrics relies, is influenced by many factors.Among them, left ventricular geometry and contractility, so low flow states may be observed in patients with either reduced left ventricular ejection fraction (LVEF) -classical low-flow low-gradient AS, but paradoxically also in patients with preserved LVEF -known as paradoxical low-flow low-gradient AS.The latter is oftentimes seen with proportionally smaller left ventricular cavities and high afterload due to arterial stiffness or hypertension.Other factors that have been shown to create a low flow state are atrial fibrillation, right ventricular dysfunction, and mitral or tricuspid valve incompetence [22,23].Dobutamine stress echocardiography is currently recommended for patients with low-flow, low-gradient AS with impaired left ventricular ejection fraction to truly severe obstruction (with afterload mismatch) from pseudo-severe AS (primary myocardial dysfunction with coexisting moderate stenosis) [13•, 16•].Patients with truly severe AS will, at an adequate hemodynamic response (increase in stroke volume by ≥ 20%), demonstrate a fixed severe AVA with a significant increase in transvalvular velocity and mean gradient.Dobutamine echocardiography is not recommended when flow is reduced but ejection fraction is preserved, and ancillary testing (see below) is advised in these cases.
"Normal-flow, low-gradient AS" represents another subset of patients who may be classified as either severe or moderate disease, depending on which criterion is used.The characterization of this group is not straightforward, and it remains unclear whether this cohort truly has a less advanced stage of disease (moderate AS), for whom surveillance may be appropriate [24], or potentially require earlier intervention.Recent meta-data indicate that these patients might benefit from valve intervention [21,25].This classification challenge can be addressed by using Q instead of SVi, because Q accounts for both ventricular function and the hemodynamic load imposed by both the stenosed aortic valve and arterial circulation.AVA is dependent not only on the compliance of the valve leaflets but also on flow rate, affecting leaflet excursion during systole.Consequently, low flow rates may lead to incomplete valve opening and subsequent overestimation of disease severity when AVA is calculated [26].Among the three conventional metrics, AVA has been suggested as the most important and sensitive independent prognostic discriminator for severe AS [21]; however, recent research has demonstrated that the prognostic value of AVA is directly dependent on Q.This suggests that Q can validate whether the calculated AVA truly represents severe disease [27••].Moreover, Q can reconcile the challenges in classifying the low-gradient population most of whom exhibit low Q [27••].As such, Q can reclassify almost all patients (> 90%) thought to have normal-flow low-gradient AS (by standard volumetric criteria; SVi < 35mls/ m 2 ) as true low-flow low-gradient AS [28].

Ancillary metrics of severity
Numerous supplementary metrics have been identified to improve classification accuracy (Fig. 1).For example, energy loss index (ELI) can be readily calculated with metrics obtained from standard echocardiography and has demonstrated superiority to AVAi in predicting adverse clinical events, such as death and hospitalization, in patients with moderate-to-severe AS, independent of mean aortic gradient and peak aortic velocity [29,30].A value of ≤ 0.52cm 2 /m 2 is consistent with severe AS [31,32].Fusion ELI, which integrates CT planimetered areas for the LVOT and aorta at the sinotubular junction, has been demonstrated to downgrade 43% of patients with severe AS by conventional AVAi (< 0.6cm 2 /m 2 ) to moderate [33].In patients with "low-gradient" severe AS (AVAi < 0.6cm 2 and mean gradient < 40 mmHg), nearly 40% were reclassified as moderate after calculation of ELI [34].These reclassified patients demonstrated a twofold reduction in the risk of the combined outcome of aortic valve replacement or cardiac mortality.Aortic valve calcification is another strong predictor of clinical outcome, and a useful adjunct metric to help confirm the presence of significant AS, particularly in patients with discordant echocardiographic indices [13•].Thresholds for confirming severe AS are 1300HU in women and 2000HU in men [35].Women tend to display greater burden of fibrosis relative to calcification compared with men.To mitigate the error introduced to the continuity equation owing to the false assumption of left ventricular outflow tract circularity, the dimensionless index is another useful metric to clarify severity.As the name implies, the index is independent of LVOT dimensions and is a ratio of the LVOT to aortic velocities -a threshold of ≤ 0.25 is consistent with severe disease (corresponding to an orifice 25% the size of what might be expected under normal flow conditions) [13•].

Left ventricular remodelling and cardiac damage
LVEF is a crucial echocardiographic measure of left ventricular systolic function and an essential tool in clinical decision-making, particularly for asymptomatic patients.For patients with severe AS, aortic valve replacement is recommended in those with LVEF < 50%, irrespective of symptoms [13•, 16•].This recommendation stems from the observation that reduced LVEF reflects more advanced disease and a decline in LVEF below normal in patients with AS heralds poorer survival [36].However, LVEF may be preserved despite prognostically adverse cellular changes within the myocardium [37,38].This may partly explain why patients with LVEF at the lower end of the normal range (50-55%) also demonstrate reduced post-operative survival, suggesting the LVEF may not be the most sensitive indicator of adverse remodelling [5,[39][40][41].A recent alternative AS classification system, comprising five stages representing various levels of upstream cardiac damage, was applied to the 1661 participants of the PARTNER 2A and 2B trials, 84.4% of whom were classified as stage 2 [42 ••].A significant correlation was observed between advancing stage and rehospitalization and all-cause death.The incorporation of left ventricular global longitudinal strain (GLS), a more sensitive marker of subtle change in left ventricular function, added incremental improvement to the system's discriminatory power [43].The addition of multi-chamber strain has also demonstrated enhancement to this classification system [44].

Artificial intelligence to improve disease classification
Advances in computer science and the exponential growth of artificial intelligence (AI) have facilitated deeper analysis of imaging and clinical datasets to explore features of prognostic significance.The scope of use for machine learning algorithms is vast.Convolutional neural networks have been employed to achieve fully automated echocardiogram interpretation and disease identification [45].Supervised and unsupervised techniques can be implemented to explore phenotypic relationships in patients with AS and even identify features associated with disease progression and clinical outcome [46, 47, 48•].A recent phenotyping study using topological data analysis revealed that different grades of AS severity, determined by classic hemodynamic criteria, occur along a phenotypic continuum with substantial inter-grade overlap [49].Although significant multidimensional overlap in echocardiographic features across the classic severity classes highlights the inherent complexity in recommending ideal intervention timing, machine learning approaches should facilitate the development of more individualized treatment decision tools.

Valve intervention
Until recently, surgical aortic valve replacement (SAVR) has been recognized as the gold standard therapy for severe AS.However, the advent of transcatheter aortic valve replacement (TAVR) introduced significant changes in terms of patient risk assessment and treatment eligibility.With the widespread adoption of this approach, TAVR now surpasses the rates of open SAVR [50], and the median age of patients undergoing TAVR has been reduced to < 75 years [51••].TAVR obviates the need for open surgery and circulatory bypass and has completely altered the landscape of procedural risk for patients with severe AS.During the nascent period of this technology, TAVR was primarily studied in patients with severe AS at high, or prohibitive, risk for open cardiac surgery as an alternative to palliation [52].The results of landmark randomized trials demonstrated clear superiority of TAVR over medical therapy or SAVR in patients at high or prohibitive surgical risk [53][54][55][56].More recently, the eligibility for TAVR has been extended to patients at intermediate [57][58][59] and even low surgical risk [60][61][62] with contemporary 5-year follow-up data demonstrating non-inferiority to open surgery [62,63].The growing experience with TAVR has led to a reduction in procedural time, complications and reduced hospital length of stay, with some centres even reporting safety and feasibility of same-day discharge [64].ACC/AHA guidelines recommend TAVR for all patients > 80 years of age or for younger patients with < 10 years expected survival, and SAVR for patients < 65 years of age (class I).For patients between 65 and 80 years of age, either modality is appropriate depending on numerous factors, highlighted in Fig. 2, in conjunction with shared decision-making.European guidelines recommend SAVR for patients aged < 75 years with an STS score < 4%, and TAVR for those ≥ 75 with an STS > 8% [16•].

Patient assessment
Key to optimizing patient outcomes is the determination of therapy eligibility, followed by appropriate procedural selection and planning.Accordingly, assessment for valve intervention requires comprehensive understanding of patient constitution and physiology.The Society of Thoracic Surgery (STS)-Predicted Risk of Mortality (PROM) score is useful for quantifying open surgical risk; however, several other factors warrant consideration, including frailty, functional status, clinical comorbidities, need for anticoagulation, prosthesis durability, aortic valve and root anatomy, and patient values and preferences (Fig. 2).Multi-disciplinary heart team discussions are crucial to synthesize these complex data and generate a tailored recommendation for patients.

Procedural planning
CT imaging plays a crucial role in procedural planning and eligibility determination.CT facilitates detailed assessment of the aortic valve morphology and leaflet calcification distribution, as well as root dimensions, coronary heights and left ventricular outflow tract geometry.These indices have an important role in guiding prosthesis selection, sizing, and implant geometry.In patients where intravenous contrast administration is contraindicated, most of these metrics can be determined using 3-dimensonal transesophageal echocardiography.Another important application of CT imaging is delineation of peripheral access suitability.Since the first in human TAVR procedure, which was performed via an antegrade femoral venous approach with transseptal catheterization [65], the retrograde femoral arterial approach has become the dominant and more simplified strategy.For the 10-15% of patients with hostile and prohibitive iliofemoral anatomy, alternate access sites such as the carotid, axillary and subclavian arteries [66], and even transcaval access [67], have been utilized with success but come at the cost of increased procedural risk.However, the uptake of iliofemoral plaque modification with intravascular lithotripsy has enabled the vast majority of contemporary TAVR procedures to be performed via the transfemoral approach [68 •].

Lifetime planning
Given the extension of TAVR eligibility to lower risk patients, the age of intervention is decreasing [51••].Therefore, evaluating the feasibility of future aortic valve intervention has become very important, owing to the known limited durability of bioprosthetic valves [69].Isolated redo surgery for degenerated surgical bioprostheses is not uncommon, and transcatheter valve-in-valve SAVR has become an established alternative for patients at high surgical risk [70].This trend has prompted the development of surgical bioprostheses optimized for future valve-in-valve procedures.The expected rise in patients who will experience structural TAVR valve deterioration warrants consideration.Valve-in-valve TAVR has become an important option for these patients [71] and should be factored into the decision regarding initial procedure and device selection.An important contributor to structural valve deterioration and poorer post-operative prognosis is patient-prosthesis mismatch (PPM), where the effective orifice area of the normally functioning bioprostheses is too small for the patient's body size [72].This is of particular concern for patients undergoing valve-in-valve procedures.TAVR with supra-annular selfexpanding valves appears to have the lowest incidence of PPM, particularly in patients with small annuli [73], and aortic root enlargement is an important consideration for patients undergoing initial SAVR.Moreover, several issues specific to TAVR warrant attention in younger patients.Paravalvular leak, which is rare with SAVR, has an association with poor post-TAVR outcomes [74].Newer generations of transcatheter prostheses have sought to address this issue.Although conduction disturbance and the need for pacemaker implantation (which is associated with poorer clinical outcomes [75]) was one of the dominant complications of TAVR during its early period, newer techniques to achieve higher implant depth (minimizing trauma to the membranous septum) have significantly reduced the need for pacemaker implantation [76,77].One of the disadvantages of higher implant depth, particularly for self-expanding valves, is the subsequent difficulty this creates with respect to coronary re-access.This is of particular importance given a significant majority of patients with severe AS will demonstrate concomitant coronary artery disease [78].As such, optimizing future coronary access requires diligent procedural planning (commissural alignment), rigorous understanding of prosthesis geometry (frame height; cell aperture), and scrutiny of preprocedural CT imaging (sinus dimensions, leaflet length, calcium burden).

Conclusions
Aortic stenosis is a complex disease that is challenging to accurately diagnose and classify.The introduction of novel imaging techniques and advanced machine learning-based approaches to large datasets continues to deepen our understanding of the disease and its trajectory, and has enabled more precise treatment recommendations, tailored to the individual.Further research is required to determine the prognostic benefit, if any, of utilizing novel classification schemes for treatment eligibility.The question remains as to whether intervention at an earlier disease stage will improve overall mortality.Transcatheter aortic valve replacement has emerged as the dominant therapeutic option for this condition.Factors such as prosthesis design, implant techniques and pre-procedural risk stratification continue to evolve in the pursuit of improving clinical outcomes.has no conflict of interest.David Muller serves as a Consultant to Medtronic, Abbott and Edwards LifeSciences.

Human and Animal Rights and Informed Consent
This article does not contain any studies with human or animal subjects performed by any of the authors.

Open Access
This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made.The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material.If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.To view a copy of this licence, visit http:// creat iveco mmons.org/ licen ses/ by/4.0/.

Fig. 1
Fig. 1 Echocardiography forms the cornerstone of AS diagnosis.The diagnostic thresholds for the key hemodynamic criteria (red boxes) are presented together with ancillary metrics (blue boxes) to assist classification.Created with BioRe nder.com.

Fig. 2
Fig. 2 Clinical, anatomic and procedural factors to consider when deciding between TAVR versus SAVR.Created with BioRe nder.com.