Low-Gradient Aortic Stenosis Outcomes
Abstract & Commentary
By Michael H. Crawford, MD, Editor
SourceS: Ozkan A, et al. Impact of aortic valve replacement on outcome of symptomatic patients with severe aortic stenosis with low gradient and preserved left ventricular ejection fraction. Circulation 2013;128:622-631. Wiegers SE. Symptomatic low-gradient severe aortic stenosis with preserved left ventricular ejection fraction: Now less of a clinical conundrum. Circulation 2013;128:576-578.
In severe aortic stenosis (AS) with symptoms, aortic valve replacement is recommended. Since the symptoms associated with AS are non-specific, when there is a discrepancy between echocardiographically calculated aortic valve area (AVA) and the gradient across the valve, treatment decisions become difficult. When valve area suggests severe AS, but the gradient is < 40 mmHg and left ventricular ejection fraction (LVEF) is normal, aortic valve replacement (AVR) may be delayed since some prosthetic valves have gradients in this range. Thus, these investigators from the Cleveland Clinic conducted a prospective observational study of 1588 symptomatic patients with isolated severe AS (AVA index < 0.6 cm2/m2) of which 260 (16%) had preserved LVEF (> 50%) and a mean gradient < 40 mmHg. All 260 patients with symptomatic low-gradient severe AS (LGSAS) underwent a complete echocardiographic examination and were followed for a mean of 28 months. Those treated medically or by balloon valvuloplasty were considered the standard therapy group. The others were treated by AVR per physician preference. The primary endpoint was all-cause mortality. AVR was performed in 123 of the 260 patients with symptomatic LGSAS. The two treatment groups were not perfectly matched in terms of baseline characteristics. As expected, the AVR patients were generally lower-risk patients. Thus, a propensity score model was used to compare the independent effect of AVR on outcome. AVR consisted of surgery (SAVR) in 94 and transcatheter (TAVR) in 29 patients. Despite a normal EF, medical patients had other measures suggesting systolic and diastolic dysfunction. Overall, 125 patients had normal stroke volume indexes (SVI) and 135 had low SVIs, which was more prevalent in medically treated patients (59% vs 44%, P < 0.02). The low SVI patients had somewhat lower EFs (59% vs 62%, P < 0.0001). During the average 28-month follow-up, 105 patients (40%) died, 32 in the AVR group and 73 in the medical group, resulting in a hazard ratio of 0.54 (95% confidence interval, 0.32-0.94; P < 0.001) for AVR. LV SVI did not influence the survival rate in the AVR patients or the medically treated patients. The authors concluded that AVR is associated with a higher 2-year survival than medical therapy in symptomatic patients with LGSAS and normal LVEF.
The development of TAVR has focused new light on the hemodynamic characterization of AS. We have known for some time that some patients with severe AS can have low transvalvular gradients. This so-called LGSAS was originally thought to be a measurement error or due to reduced LV function, which could be sorted out using dobutamine stress echo to improve LV function. More recently, a subgroup of LGSAS has been identified with normal LVEF, which is unexpected since normal LV function should produce a robust gradient in severe AS. Several papers, including this one, have now described two subgroups of LGSAS with normal LVEF patients: those with normal stroke volume and those with reduced stroke volume. The latter group has been called "paradoxical low-flow, low-gradient AS" because the low stroke volume is unexpected with a normal EF. This is confusing because a low gradient with a normal EF itself seems paradoxical. So, I hope this term does not survive.
What does all this mean? My answer is that EF is a crude measure of LV performance that is load dependent. Depending on your LV diastolic volume, a normal EF can be associated with a normal SV or a reduced SV. This study and more sophisticated studies have shown decrements in LV performance measures beyond EF in AS patients, and biopsies in humans have shown increased fibrosis that would tend to impair diastolic function. Thus, in the case of severe AS, EF is not a particularly useful measurement.
Clinically, the patients who had AVR after adjustment for known confounders did better than those treated medically. Remember, all of these patients were symptomatic. So, if you are confident that your echocardiographic estimation of valve area is correct and the patient has a normal EF and symptoms, then it doesn’t really matter what the gradient is. Such patients did much better with valve replacement in this prospective study and previous retrospective studies. However, several aspects of this study have probably magnified the differences between medical and surgical therapy. First, a propensity score model cannot account for all variables that influence prognosis. The surgical patients were clearly healthier based on baseline characteristics. Also, important patient characteristics such as frailty and dementia, which may have affected the decision not to perform surgery, were not considered. In addition, about 25% of the patients had TAVR, which will lower the surgical group’s mortality since there were no deaths within 30 days of TAVR. Despite these factors that undoubtedly increased the apparent benefit of AVR, a two-fold difference in mortality between the two groups is impressive and should be considered in symptomatic patients because they are more likely to feel better after AVR as well. This study should not be applied to asymptomatic patients and patients with moderate AS or reduced EF.