The published correlation of Doppler versus catheterization mean gradients averages 0.90,^3,⁴ but importantly, 1 standard error of estimate (SEE) of Doppler gradient versus catheterization gradient is actually 10 mm Hg [SEE range, 6–12 mm Hg].⁵

Rahimtoola ⁵ emphasized consideration of the SEE by Doppler when describing AS severity (by gradient):

• Highly likely severe: mean gradient >70 mm Hg

• Probably severe: mean gradient >60 and <70 mm Hg

• Uncertain if severe: mean gradient <50 mm Hg

Recall that the pressure recovery phenomenon at aortic root level is seen with small aortic roots (dimensions <30 mm) and restrictive planar divergent orifices. Pressure recovery may add 15% to 30% to the gradient.

LVSWL, originally a catheterization-based parameter, can be approximated by echocardiography, and has been validated as a way to discriminate clinical end-points: an LVSWL ≤25 is the best predictor.⁶ The concept expresses kinetic energy loss across the aortic valve in reference to the aortic pressure.

Continuity Equation–Derived Aortic Valve Area Issues

The idea of aortic valve area as a conceptual expression of AS is attractive, because it is intended to be a flow-independent description of AS severity, and it is widely accepted that flow does vary across stenotic aortic valves, reducing the predictiveness of gradient alone. Because flow across the aortic valve per cardiac cycle may vary, the gradient should be offered in the context of the stroke volume (index). Unfortunately, precise determination of the stenotic aortic valve area, by echo and also by catheterization, is neither as simple nor accurate as is believed. The calculation of AVA by echo and by catheterization actually is not “flow-independent” by either technique, because so many parameters are involved in the equations that the total introduced error becomes significant and sometimes problematic. Furthermore, plastic deformation of the aortic valve (i.e., greater opening under higher gradients and less opening under lower gradients) is known to occur.

Place less emphasis on the valve area alone; instead, emphasize the combination of area, gradient, and stroke volume. Unless these afford a congruent assessment of aortic stenosis severity, further consideration is needed.

Continuity-derived AVA is not flow independent.⁷

The multiple variables used in the calculation of AVA may allow for introduction of considerable composite error.

The average published correlation of Doppler-derived AVA versus catheterization-derived AVA is r = 0.90. However, recall that the SEE of Doppler AVA versus catheterization gradients is, on average, 0.3 mm Hg [0.1–0.4 cm²].⁵

Rahimtoola ⁵ suggests consideration of patient body size indexing of the AVA.

• Severe: <0.3 cm²/m²

• Moderate: 0.3–0.59 cm²/m²

• Mild: >0.6 cm²/m²

Report AVA only to a single decimal place. A calculation accurate at best to ± 0.3 cm² cannot be realistically described to a second decimal place.

Peak Systolic Velocity

Calculations of gradient and area entail squaring and multiplying (4 × V², 4 × V_mean², 0.785 × D² × V₁/V₂), thereby conferring additive error to their calculation. Consequently, the case has been made in favor of measurement of the peak velocity alone, avoiding the error accrued from calculation of gradient and area. Peak velocity ≥ 4.5 m/sec is a reliable descriptor of severe AS, in the absence of noncardiac factors that may aggravate velocity. As a parameter, it is clinically useful, and as an example it serves as a reminder that calculated variables may diminish their own worth by the error accrued in yielding them. Peak systolic velocity is a robust means of describing the severity of AS.

Summary of Aortic Stenosis Descriptors

The important task is to determine which cases of AS are hemodynamically severe, and may benefit from surgery. There is little worth in debating moderate versus mild; both are in effect “nonsevere” and would not independently merit intervention. One significant exception is that AVR usually is performed in the scenario of CAD to undergo aortocoronary bypass grafting, in the presence of moderate AS (mean gradient ≥ 25 mm Hg), although the threshold for performing AVR in this scenario remains vexingly unclear.

Possibly severe in the presence of a normal stroke volume

• AVA < 1 cm² (for a person of average body size) or

• Indexed AVA < 0.5cm²/m² and a mean gradient ≥ 50 mm Hg

Severe in the presence of a normal stroke volume

• AVA < 0.75 cm² (for a person of average body size) or

• AVA < 0.3cm²/m² and a mean gradient ≥ 50 mm Hg

Other Issues

Although the focus of an echocardiographic study is on detailed examination and description of the valve and related hemodynamics, it also is necessary to assess for associations, complications, and concurrent diseases.

If the aortic valve is bicuspid, describe possible associations. Recall that a high-velocity AS jet may carry around the aortic arch—in other words, higher-velocity flow in the proximal descending aorta may not be coarctation generated, and represents only transmission of the AS jet. The diagnosis of coarctation of the aorta by echocardiograophy requires demonstration of a focal step-up of flow velocity.

Describe the dimensions and appearance of the root and ascending aorta. These are potentially important surgical details that may modify the approach to surgery. Furthermore, if a case is obviously severe, and the aortic annulus is small, the size of the annulus should be mentioned. An unusually small root will allow only a small prosthesis, which engenders a gradient little better than the AS it was supposed to relieve—an outcome known as “patient–prosthesis mis-match.” Although such a mismatch probably is not as significant a clinical problem as has been purported, it still has validity in some cases.

The presence of LV hypertrophy is expected. Its absence is conspicuously inconsistent with severe AS. The presence of wall motion abnormalities is important.

Concurrent Subvalvar Stenosis

Subvalvar AS may masquerade as valvar AS, or may be an unsuspected concurrent lesion. Concurrent subvalvar AS confounds accurate estimation of the aortic valve hemodynamics, and is not treated by AVR alone.

Subvalvar AS is revealed by the PISA (flow convergence) at a level well beneath the aortic valve (>1 cm beneath, but influenced by the Nyquist limit). Formerly, emphasis of detection was on pulsed wave Doppler sampling; however, as color flow mapping is essentially a flow map of pulsed-wave Doppler, clear color Doppler depiction of PISA well away from the valve is equivalent.

A two-fold focal step-up in velocity and a “dagger” or tooth shape of the profile are consistent with a intracavitary stenosis.

Provocative maneuvers (e.g., posture change, Valsalva maneuver, leg raise) may reveal the subvalvar gradient to be dynamic.

Zoom views are helpful to localize the step-up, and begin to resolve its nature, e.g., muscular, membrane, or tunnel, or from multiple causes.

TEE is superbly able to define the anatomy responsible for subvalvar obstruction; cardiac MRI is very good, and cardiac CT likely is as well.

Although one catheter study has suggested a 30% incidence of subvalvar gradients,⁸ the incidence does not appear to be as high as that.

In the presence of subvalvar AS, the routine sampling of V₁ will fall into the flow convergence/flow acceleration or vena contracta of subvalvar stenosis. If the subvalvar velocity is >1.5 m/sec, the calculations of aortic valve area are skewed.

Low-Gradient Severe Aortic Stenosis

The definition of “low-gradient severe AS” is unresolved and variable, but may be approached as follows:

• Severe AS with an AVA ≤ 0.7 cm² with a mean gradient of ≤30 mm Hg due to reduced stroke volume.⁹ Several papers have used an AVA of ≤1 cm² and a mean gradient of ≤40 mm Hg.¹⁰

The challenge is to distinguish patients with end-stage AS with failing LV systolic function (responsible for the low gradient) from patients with moderate AS with poor LV systolic function, in whom a factor other than AS is responsible for the low gradient.

Low-gradient severe AS can be discounted by a convincingly normal stroke volume: if the gradient is low and the stroke volume is normal, then low-gradient severe AS is unlikely to be present.

Citing low EF% or “grade” to describe low-output AS is not adequate. The stroke volume should be shown to be low to establish that the per-beat output is low, and potentially responsible for the low gradient. There may be normal stroke volume with a low EF% LV (if dilated) and low stroke volume with an LV with normal EF% (if it is under-loaded or if there is severe MR). Therefore, LV grade is not synonymous with output.

EF% generally increases after AVR ^10,¹¹ unless a perioperative infarction occurs, or the LV is intractably stiff. It appears that selected cases of low-gradient severe AS still benefit from AVR,¹² although the data in this field are preliminary.¹³

In the presence of low output, both the Gorlin catheterization formula^11,14 and continuity methods are less accurate in the estimation of AVA.

When catheterizing suspected low-gradient AS, use of either two catheters or double-lumen catheters should be considered to eliminate phase shift artifacts, which can introduce a difference of >10 mm Hg in the gradient recording from a femoral side-arm.

Low-gradient AS should be reported as a possibility or probability, depending on how strongly it is believed to be present.

Valve resistance has been proposed as a means of establishing the severity of AS (>300 dynes/sec/cm⁻⁵ = severe disease) that is independent of flow output, but it has not been proved to be better than valve area calculation,⁹ nor has it been popular. It appears less useful than LVSWL.⁶

Determination of contractile reserve in low-gradient severe AS cases

• Consideration of pharmacologic stress (dobutamine up to 20 μg/min/m²)

• Determination of AVG and AVA in suspected cases of low-gradient AS (exercise or dobutamine) via echocardiography or catheterization to increase stroke volume by more than 20% and cardiac output (CO) to 4.5 L/min in non-coronary cases is reasonable. The gradient changes can be interpreted in the context of the change of the stroke volume across the valve.

• If stroke volume increases and the AVA increases outside the severe range, then the cause of the low gradient was low output (i.e., pseudostenosis), but the AS is not severe.

• If the stroke volume increases but the AVA remains within the severe range, then the AS is severe (and pseudostenosis is not present).

• However, AVR reduces mortality whether or not there is contractile reserve, and medical therapy is a poor option whether or not there is contractile reserve. Therefore, provocative testing should be done, although it has not been unequivocally proven to be helpful.^9,¹³

• Rahimtoola ¹⁵ would caution that changes in stroke volume are labeled as an index of contractility and as the hemodynamic responses to dobutamine are complex.