Aortic stenosis (AS) is the most common valve disease in Europe, and is most often due to degenerative etiology.1 It is estimated to affect 2-7% of the population aged over 65 years.2 Degenerative AS progresses slowly and is associated with various clinical manifestations including angina, syncope, cardiac rhythm disturbances, heart failure and sudden death.1

According to the European and American guidelines on valvular disease, severe AS is defined as aortic valve area (AVA) of ≤1 cm2, mean pressure gradient (MPG) between aorta and left ventricle of ≥40 mmHg, or peak flow velocity (PFV) of ≥4 m/s, with normal cardiac output.1,3

In the absence of valve replacement, AS evolves to a form of heart failure with left ventricular (LV) dilatation and dysfunction. In this stage there is a reduction in transvalvular gradients, caused by diminished systolic flow through the aortic valve due to impaired LV systolic function. This entity, first described by Carabello et al. in 1980,4 is termed low-flow, low-gradient aortic stenosis (LF-LG-AS). It is characterized by an AVA of ≤1 cm2 or indexed AVA (AVAi) of ≤0.6 cm2/m2, MPG of <40 mmHg and left ventricular ejection fraction (LVEF) of ≤40%. Patients with LF-LG-AS account for only 5-10% of those with severe AS.5–7

In 2007, Hachicha et al.8 described for the first time a form of AS with reduced systolic flow but, paradoxically, preserved LVEF. This new entity, which they termed paradoxical low-flow, low-gradient aortic stenosis, was defined as AVA of ≤1 cm2, MPG of <40 mmHg, and stroke volume index (SVI) of ≤35 ml/m2, despite LVEF of ≤50%.3,5,9 This entity has been shown to have a significant prevalence and it is thought that it may represent an advanced form of AS.8,10

The aim of the present article was to perform a systematic review of the literature on paradoxical AS, discussing aspects of definition, clinical characteristics, diagnosis, prognosis, natural history, and treatment.

A search was carried out in the PubMed database up to December 29, 2014, using the following search terms: “Paradoxical aortic valve stenosis” and one of: “low flow”, “low gradient”, “assessment”, “left ventricular ejection fraction”, “echocardiography”, “treatment”, or “prognosis”. The search covered articles in English published since 2007 and included original articles, clinical trials, reviews and animal trials.

The filtered search yielded 61 articles, of which 23 were rejected on the basis of the title and abstract as being unrelated to the study aim. A complete reading was performed of the remaining 38, of which 13 were excluded from the present review: six letters to the editor, two case reports and one clinical vignette, one published only as an abstract, two that did not in fact refer to paradoxical AS, and one editorial that was considered irrelevant. In order not to miss any important articles, the reference lists of the selected articles were perused and 32 more publications were added to the original list. The guidelines on valvular disease of the American and European societies of cardiology were also consulted. The final list consisted of 57 articles (Figure 1).

Figure 1.

Flowchart of the study design.

(0.2MB).

The quality of the main studies analyzed in the present review was assessed (Supplementary Table 1) by an external consultant using the Effective Public Health Practice Project tool (http://www.ephpp.ca/tools.html).

The search results produced 14 original articles on paradoxical AS, of which four were retrospective (6,8,54,55) and seven were prospective (11,16,17,22,41,52,53). Two substudies of randomized trials (49,51) and one of a multicenter registry (56) were also included.

The aim of around a third of the studies was to compare paradoxical AS with the classical AS phenotype (6,8,11,22,51). One study included a comparison with patients with moderate AS (8) and another with patients with LF-LG-AS (52). Two assessed the interaction between transvalvular gradient and SVI in patients with AVA of ≤1 cm2 and LVEF of ≤50% (16,17). Six focused on surgical treatment (6,8,16,17,22,53) and three on percutaneous treatment of patients with paradoxical AS (49,55,56). These 14 articles are summarized in Table 1.

Paradoxical AS is defined as AVA of ≤1 cm2, MPG of <40 mmHg, and SVI of ≤35 ml/m2, despite LVEF of ≤50% (Figure 1).8 This condition is usually the result of marked LV concentric remodeling with interstitial myocardial fibrosis, and reduction of the ventricular chamber and its compliance.8,11–15 There is reduced longitudinal myocardial strain in the phenotype of paradoxical AS despite good global LV systolic function.6

Low systolic flow hampers assessment of the severity of AS, since the gradient is proportional to the fourth power of the flow, and any reduction in flow leads to a large reduction in gradient and hence underestimation of disease severity.9 Furthermore, transvalvular flow rate depends not only on ejection volume but also on ejection time.19 Other factors such as atrial fibrillation (AF), mitral regurgitation, mitral stenosis and tricuspid regurgitation can also play an important part in reducing ejection volume and thus contribute to situations of low flow and low gradient.7,16

Estimates of the prevalence of paradoxical AS range between 3% and 35% of patients with severe degenerative AS (Figure 2).6,7 This discrepancy may be due to the characteristics of study populations.8,16 For example, in an echocardiographic study of 1704 consecutive patients with severe AS and preserved LVEF (≤50%), Eleid et al.7 found a prevalence of paradoxical AS of only 3%, whereas in another echocardiographic study, of 397 patients with AS, Kusunose et al.6 documented a prevalence of paradoxical AS of 35%. Another study, on a sample of patients with severe AS referred for cardiac catheterization,17 observed a prevalence of 26%, a similar figure to echocardiographic studies.

Figure 2.

Prevalence of paradoxical aortic stenosis as a percentage of all patients with severe aortic stenosis. cath: catheterization; echo: echocardiography.

(0.24MB).

Paradoxical AS appears to be more frequent in females and older patients. Compared to classical severe high-gradient AS, more patients with paradoxical AS have a history of hypertension, diabetes, AF and coronary artery disease.6,7,9,10,17,18 Hypertension is common in patients with AS, but blood pressure (BP) in those with paradoxical AS may be pseudonormalized by the low-flow pattern associated with increased vascular stiffness.8 Its symptomatology is similar to that of classical severe AS, although its onset is more insidious.1,10

When paradoxical AS is suspected, as well as PFV, aortic transvalvular gradients and AVA, assessment should include vascular load, global LV hemodynamic load, LV function beyond LVEF, LV geometry, and myocardial damage.19

Transthoracic echocardiography is the method of choice for hemodynamic study in AS. Cardiac catheterization is not routinely indicated except when echocardiography is not diagnostic or when there are discrepancies in clinical findings.20,21

Problems with calculation of aortic valve area

AVA is calculated by the continuity equation, which is based on the principle of conservation of mass, assuming that the volume that passes through a given area upstream of the aortic valve is the same as that which passes through the valve. This measure is less flow-dependent than other parameters such as velocity and gradient, but is susceptible to error, since it is only an indirect measure of valve area. Firstly, the continuity equation measures the size of the functional orifice instead of that of the anatomic orifice and thus neglects the coefficient of orifice contraction, and so under normal flow conditions underestimation of the anatomic orifice by the continuity equation is expected to be around 10-15%.22 Furthermore, LV ejection volume (the numerator in the continuity equation) is usually calculated by two-dimensional echocardiography on the basis of cross-sectional area, assuming the left ventricular outflow tract (LVOT) to be circular. However, Doddamani et al. used multislice computed tomography (CT)23 and three-dimensional echocardiography24 to show that the LVOT is in fact more often elliptical than circular. This can result in AVA being underestimated by a further 15%.24

Since the severity of AS can be under- or overestimated if the patient has a large or small body surface area (BSA), respectively, it is essential to index valve area to BSA, yielding AVAi, although it should be noted that such indexing may not be appropriate at extreme values of BSA.1

Consistency between echocardiographic criteria of AS severity (AVA by the continuity equation, PFV and MPG) was established following an analysis of 3482 echocardiograms of patients with good global LV systolic function and AVA <2 cm2. It was shown that the criterion of AVA ≤1 cm2 overestimated severity compared to the other two criteria (PFV ≤4 m/s and MPG ≤40 mmHg). The authors accordingly suggested that the cutoff of AVA to define severe AS should be <0.8 cm2, which more often corresponds to PFV of ≤4 m/s and MPG of ≤40 mmHg. A limitation was that the authors did not present the clinical results of this reclassification.25

To summarize, although AVA is an essential parameter in the stratification of AS severity, there are significant limitations to its measurement.

Three-dimensional echocardiography is an alternative method to measuring AVA that has been shown to be more accurate in assessing AVA and AS severity than volumetric methods such as the continuity equation by two-dimensional echocardiography.26

Projected AVA (AVAproj) is another way to estimate AVA in situations of low systolic flow and associated general systolic dysfunction. It is based on a normal systolic flow (250 ml/ms) by the formula

where VC is vascular compliance and Qrest is systolic ejection flow at rest. AVAproj, which is obtained during dobutamine stress echocardiography, appears to offer a good alternative to standard echocardiographic parameters.27

New parameters for assessment of AS have appeared that are improving understanding of the disease.

The degree of aortic valve calcification is a strong predictor of disease progression.19 Although it can be semi-quantitatively assessed by echocardiography, a more precise quantification of calcification is provided by CT.19 According to the American guidelines, a calcium score of >1000 Agatston units identifies severe aortic valve calcification.3 This method is not affected by the patient's hemodynamic characteristics and is thus useful in low-flow states such as paradoxical AS.39 Messika-Zeitoun et al.40 demonstrated a strong association between aortic valve calcification and AVA (r=-0.79, p<0.01).

In a study of 250 consecutive patients treated by SAVR, Clavel et al.41 found that in 70% of patients with paradoxical AS, stenosis was considered severe on the basis of the weights of valves excised during surgery. This percentage is similar to that found in patients with classical severe high-gradient AS (86%).41 This study confirms that a large proportion of patients with paradoxical AS in fact have a severe form of the disease and highlights the limitations of the echocardiographic methods used for stratifying AS patients, especially valve gradients.41

Current guidelines do not differentiate between methods for measuring AVA (catheterization or Doppler echocardiography). AVA can be assessed invasively using the Gorlin formula,20 but this is dependent on systolic flow and is unsuitable in low cardiac output states. Doppler echocardiography measures the pressure gradient through the valve based on peak systolic flow at the vena contracta, while catheterization determines the pressure gradient between the left ventricle and the ascending aorta, several cm from the valve.19 The total energy of aortic flow is the sum of kinetic and static energy (pressure). Flow convergence through the stenotic valve converts static to kinetic energy, reducing pressure at the vena contracta. When velocity falls distal to the stenosis, part of the energy is converted back to static energy due to the phenomenon of pressure recovery.42 The clinical implication of this phenomenon is that Doppler echocardiography underestimates AVA compared to invasive assessment.28 To overcome this limitation, Garcia et al. developed a new index, the energy loss coefficient (ELCo),43 to adjust AVA to pressure recovery in the ascending aorta.28 ELCo is calculated by the formula

As mentioned above, AS is nowadays considered to be a systemic vascular disease, with increased systemic vascular stiffness. Recent studies have shown that the vascular mechanics of the aorta can be assessed by speckle-tracking echocardiography, an imaging indicator of vascular stiffness,45,46 and that values are lower in AS than in aortic regurgitation.45 Moreover, patients with moderate to severe low-flow AS present lower aortic strain values than those with normal systolic flow.46 Although promising, assessment of valve mechanics by speckle-tracking echocardiography is currently only a research tool due to limitations arising from the need for a good acoustic window on the aorta and the lengthy analysis required.

New markers, including neurohormones, adipokines and extracellular matrix modulators, have recently been identified as prognostic and diagnostic factors of severe low-flow, high-gradient AS with preserved LVEF.47 Their future potential for assessment of patients with paradoxical AS should be the subject of further research.

When assessment of a patient with AS reveals discordant values of AVA and MPG (AVA ≤1 cm2 and MPG <40 mmHg), and there is good LV systolic function but reduced systolic flow (SVI ≤35 ml/m2), the possibility of paradoxical AS should be considered. According to Pibarot et al.,48 a systematic approach is required (Figure 3). Other morbidities such as significant mitral or tricuspid valve disease should be ruled out at the outset, as should measurement errors and misplacement of the Doppler sample volume. Echocardiographic assessment should be performed under normal blood pressure conditions. AVA should be indexed to BSA and body mass index should also be considered. An AVAi of ≤0.6 cm2/m2 indicates severe AS. Signs of LV remodeling suggest the presence of paradoxical severe AS. When the severity of AS is in doubt, additional information can be obtained from other parameters such as aortic valve calcium score, anatomical AVA, Zva, ELCo, and strain parameters, which may indicate not only severity but also the diagnosis of paradoxical AS.19

AVA: aortic valve area; AVAi: aortic valve area indexed to body surface area; BSA: body surface area; BNP: brain natriuretic peptide; CT: computed tomography; ELCo: energy loss coefficient; GLS: global longitudinal strain; LV: left ventricular; LVEDD: left ventricular end-diastolic diameter; LVEDVi: indexed left ventricular end-diastolic volume; LVEF: left ventricular ejection fraction; LVOT: left ventricular outflow tract; MPG: mean pressure gradient between aorta and left ventricle; RWTR: relative wall thickness ratio; SVI: stroke volume index; Zva: valvuloarterial impedance.'> AVA: aortic valve area; AVAi: aortic valve area indexed to body surface area; BSA: body surface area; BNP: brain natriuretic peptide; CT: computed tomography; ELCo: energy loss coefficient; GLS: global longitudinal strain; LV: left ventricular; LVEDD: left ventricular end-diastolic diameter; LVEDVi: indexed left ventricular end-diastolic volume; LVEF: left ventricular ejection fraction; LVOT: left ventricular outflow tract; MPG: mean pressure gradient between aorta and left ventricle; RWTR: relative wall thickness ratio; SVI: stroke volume index; Zva: valvuloarterial impedance.' title='Algorithm for diagnosis of paradoxical aortic stenosis. 3D echo: three-dimensional echocardiography; AF: atrial fibrillation; AS: aortic stenosis; AVA: aortic valve area; AVAi: aortic valve area indexed to body surface area; BSA: body surface area; BNP: brain natriuretic peptide; CT: computed tomography; ELCo: energy loss coefficient; GLS: global longitudinal strain; LV: left ventricular; LVEDD: left ventricular end-diastolic diameter; LVEDVi: indexed left ventricular end-diastolic volume; LVEF: left ventricular ejection fraction; LVOT: left ventricular outflow tract; MPG: mean pressure gradient between aorta and left ventricle; RWTR: relative wall thickness ratio; SVI: stroke volume index; Zva: valvuloarterial impedance.'/>

Figure 3.

Algorithm for diagnosis of paradoxical aortic stenosis. 3D echo: three-dimensional echocardiography; AF: atrial fibrillation; AS: aortic stenosis; AVA: aortic valve area; AVAi: aortic valve area indexed to body surface area; BSA: body surface area; BNP: brain natriuretic peptide; CT: computed tomography; ELCo: energy loss coefficient; GLS: global longitudinal strain; LV: left ventricular; LVEDD: left ventricular end-diastolic diameter; LVEDVi: indexed left ventricular end-diastolic volume; LVEF: left ventricular ejection fraction; LVOT: left ventricular outflow tract; MPG: mean pressure gradient between aorta and left ventricle; RWTR: relative wall thickness ratio; SVI: stroke volume index; Zva: valvuloarterial impedance.

(0.21MB).

Adapted from Pibarot et al.48

Various studies have indicated that the combination of low systolic flow, low transvalvular gradient, and preserved global systolic function is a marker of poor outcome.

• •

  Hachica et al.,8 in a retrospective study of 512 consecutive patients with severe AS and preserved LVEF, showed that low flow (SVI ≤35 ml/m2) was associated with worse outcome at three years.

• •

  Mohty et al.,17 in a study of 768 patients with severe AS and preserved LVEF, observed that low flow and low gradient was associated with reduced long-term survival after adjustment for other factors.

• •

  In an observational study of 1704 consecutive patients with severe AS and preserved LVEF, Eleid et al.7 showed that the low-flow, low-gradient phenotype was the strongest predictor of mortality.

• •

  Lancellotti et al.,49 in a prospective study of 150 consecutive patients with severe AS, demonstrated that patients with low flow and low gradient had worse clinical outcomes than those with normal flow and low gradient; in this study both low flow and low gradient were independent predictors of poor outcome.

• •

  In a subanalysis of 971 patients with severe AS of the Placement of AoRTic TraNscathetER Valve Trial Edwards SAPIEN Transcatheter Heart Valve (PARTNER) trial, Herrmann et al.50 showed that only low flow independently predicted mortality in both cohort A (high-risk patients) and cohort B (inoperable patients). All-cause mortality at two years was also significantly higher in low-flow than in normal-flow AS.

There is thus general agreement in the literature that low-flow states in severe AS with good global systolic function are associated with worse medium- and long-term prognosis.10 Cardiovascular events are more frequent in patients with paradoxical AS than in those with normal-flow AS,13,51 and survival at one and five years in a retrospective study of 1589 patients by Clavel et al. was less than in patients with moderate AS.10

Not all authors draw the same conclusions. In a recent study,22 Maes et al. prospectively assessed 349 patients with severe AS and preserved LVEF, divided into two groups according to MPG (>40 mmHg or ≤40 mmHg). Prognosis as measured by all-cause survival at four years was better in patients with paradoxical AS than in those with classical severe high-gradient AS.

Similarly, in a subanalysis of the SEAS study, Jander et al.52 showed that patients with paradoxical AS had similar clinical outcomes to those with moderate AS.

The natural history of paradoxical AS has been the subject of intense debate since 2007, and as reported above, there seems to have been a change in the disease paradigm. As shown in Model A (Figure 4), after a stage in which AS reaches a certain level of severity, in the absence of treatment the patient may evolve to a phenotype in which LV dilatation and systolic dysfunction predominate, leading to reduced systolic flow and transvalvular gradients. The initial studies by Hachicha et al.8 and Lancellotti et al.49 suggested that patients with severe AS can also evolve to a different phenotype. In this view, paradoxical AS represents an advanced stage of aortic valve disease due to marked LV interstitial fibrosis, reduced chamber size and increased systemic vascular stiffness, leading to diminished LV systolic flow and hence reductions in pressure gradients and aortic valve area. As described above, the fact that patients with paradoxical AS have worse outcome than those with classical (high-gradient) severe AS corroborates this model.

LF-LG-AS: low-flow, low-gradient aortic stenosis.'> LF-LG-AS: low-flow, low-gradient aortic stenosis.' title='Two models of the natural history of aortic stenosis. AS: aortic stenosis; LF-LG-AS: low-flow, low-gradient aortic stenosis.'/>

Figure 4.

Two models of the natural history of aortic stenosis. AS: aortic stenosis; LF-LG-AS: low-flow, low-gradient aortic stenosis.

(0.1MB).

Recent findings, however, point to a different evolution of the disease (Model B in Figure 4). Maes et al.22 suggest that paradoxical AS represents an intermediate stage in progression to severe high-gradient AS. The prognosis of patients with paradoxical AS is similar to that of those with moderate AS and is better than in high-gradient severe forms. Furthermore, in the absence of valve replacement, severe AS evolves to a low-flow, low-gradient form and not to paradoxical AS. According to Maes et al., the concept of paradoxical AS is probably due to inaccuracies in the calculation of AVA using the continuity equation, as discussed above.

AS should not be considered an isolated disease of the aortic valve, but rather a systemic vascular disease involving atherosclerosis and arteriosclerosis. In addition, its clinical manifestations also depend on ventricular remodeling and LV function.8 Treatment should bear all these factors in mind.8

Paradoxical AS was described for the first time in 2007 and its phenotypic characteristics have been confirmed by various groups. However, questions remain concerning inconsistencies in the continuity equation for measurement of AVA that may explain the conflicting findings in this phenotype and could lead to reclassification of some forms of severe AS as moderate AS. Non-invasive diagnostic methods have been developed that can better identify the paradoxical form of severe AS, including assessment of LV remodeling and of the vascular component, measurement of VA and of the LVOT by three-dimensional methods (tomographic and non-tomographic), and calculation of the calcium score of the aortic valve.

The data presented above appear to indicate that the prognosis of patients with paradoxical AS is poor, although it is not clear whether model A or B (Figure 4) explains the stage it represents in the natural history of AS.

It should be noted that most of the studies analyzed indicate that the prognosis of patients with paradoxical AS is better following valve replacement than with medical therapy. In the European and American guidelines valve replacement is a class IIa recommendation (albeit with level of evidence C) for symptomatic paradoxical AS. The limitations of studies assessing the prognostic impact of valve replacement in patients with paradoxical AS – including patient selection, study design (single center, retrospective analysis, lack of randomization) and confounding variables – are summarized in Supplementary Table 1.

The phenotype of paradoxical AS, characterized by low systolic ejection volume and low gradient, is common in patients with AS and preserved LV systolic function. Measurement errors, inconsistencies in the continuity equation and lack of BP control can lead to contradictory findings when assessing a patient with AS. Although there is some disagreement in the literature, the prognosis of symptomatic paradoxical AS is poor and in such cases valve replacement should be considered.

The present review includes only articles published up to December 2014. It should be noted that the representativeness of the studies in the quality assessment was low (25%).

RC performed the bibliographic search, produced the literature summary and wrote the first version of the manuscript. RT designed the review, discussed the results, modified the initial document for submission and responded to the reviewers. MJV carried out the quality assessment and played an important part in the response to the reviewers. LG contributed to the discussion of the results and to correction of the manuscript.

The authors have no conflicts of interest to declare.