July 15th, 2013

The Conundrum of Low-Gradient Severe Aortic Stenosis with Preserved LVEF


CardioExchange’s John Ryan interviews Thomas Marwick about his research group’s study, published in Circulation, about how patients who have severe aortic stenosis with low gradient and a preserved LV ejection fraction fare after aortic valve replacement.


Researchers prospectively identified 260 patients with symptomatic low-gradient severe aortic stenosis and followed them for a mean of 28 months. Nearly half (47%) underwent aortic valve replacement (AVR); the rest received optimal medical therapy alone. The mortality rate was significantly lower in the AVR group than in the medical group (30% vs. 70%). AVR remained a strong predictor of survival in propensity-matched analyses. AVR’s benefit was evident regardless of stroke volume index (SVi).


Ryan: Given your findings, should the standard of care for symptomatic low-gradient aortic stenosis be aortic valve replacement, or do we need a randomized controlled trial?

Marwick: It would be great to have an RCT, but in the current environment it will be hard to get funded, even in the U.S. Maybe this will be a spinoff of TAVR trials and, hence, funded by industry. Surgical trials are obviously hard to do. For now, this matched cohort study design is likely to be as good as we can manage. I think it helps me with the low-gradient patient when the surgeon says that an AVR will not have a much lower gradient than the native valve. These patients seem to benefit from surgery.

Ryan: How can we increase physicians’ awareness of this condition?

Marwick: I think there’s still a lot of confusion about what to do when the gradient and aortic valve area (AVA) don’t match. We all know that AVA is a derived parameter, so there is a tendency to underweight the AVA. Clearly, we should check the measurements, especially of the left ventricular outflow tract (LVOT). I think 3D has a real role in proper estimation of the LVOT. But when that has been done and the AVA is small in the setting of a low gradient, we shouldn’t ignore this, especially if the valve looks ugly and the stroke volume is reduced.

Ryan: In turn, are you worried that an emphasis on this condition could lead patients who do not actually have AS to be labeled with it?

Marwick: This is a real concern, but not as much as failing to intervene because AS has not been recognized. However, like most things in valve disease, it involves sensible integration of several parameters. If the valve opens adequately and the LVOT measurement is questionable, then making a call on AVA with a low gradient would be unwise. Sometimes transesophageal echocardiography, exercise testing, or more-advanced imaging might be needed.

Ryan: Should we have a standard approach to diagnosing low-flow AS?

Marwick: Again, I think it goes back to integrating several parameters and weighing their reliability. So it’s hard to make a recipe, but here’s what I do:

  • Start with AVA, gradient, and dimensionless severity index (DSI). If they match up, it’s all good. If they don’t match, the next step is to explore why.
  • Reassess the LVOT. If the image quality is horrible and you’re worried about the accuracy of LVOT measurement, down-weight AVA and use DSI.
  • If the LV ejection fraction is poor or LV SVi is <35 mL/m2, I down-weight gradient.
  • Watch out for the severity of morphologic changes in the AV.
  • Keep an eye on the left ventricle. If LVEF is reduced, filling pressures are high, or there is unexpected LV hypertrophy (without other explanations), you might be missing an aortic valve problem.



Share your experiences with patients who have severe aortic stenosis with a low gradient and preserved LV ejection fraction. Will this study change your practice?

One Response to “The Conundrum of Low-Gradient Severe Aortic Stenosis with Preserved LVEF”

  1. I completely agree with Marwick. There are about 20-25% of AS patiente, expecially very old (more than 80ths) with history of hypertension and/or CKD in this condition: low AVA, low Grad, and LVEF >=40%.
    The tipical hallmarks are: DSI <25% (suggesting severe AS), low mean AS gradient <30mmHg and normal or near normal EF%. Generally the valve is diffusely calcified (more than in a previous echo) with hypomobile cuspids. You can help your "decision making" by an TE-echo, where in some case we can accurately measure the planimetric AVA. In some other cases also we can perform a LD-Dobutamine-echo and observe a "discrete" change of gradient from rest 30 to 50 or 60 mmHg. In this manner we can safely take the decision to send the patient to cardiac surgeon for a …better life!