Surgery for Asymptomatic Severe Aortic Stenosis

Abstract & Commentary

By Michael H. Crawford, MD

Source: Kang DH, et al. Early surgery versus conventional treatment in asymptomatic very severe aortic stenosis. Circulation. 2010;121:1502-1509.

The ACC/AHA guidelines give a class ii b recommendation for surgery in asymptomatic patients if aortic stenosis is very severe (area < 0.6 cm2) and operation risk is low. However, the outcome of this approach vs. following patients until symptoms or left ventricular dysfunction occurs has not been tested. Thus, this group from South Korea evaluated their prospective database on severe aortic stenosis patients to answer this question. They included asymptomatic patients with a valve area of ≤ 0.75 cm2 and one of the following: peak velocity ≥ 4.5 m/s or a mean gradient ≥ 50 mmHg on Doppler echo. They excluded symptomatic patients, those with an ejection fraction < 50%, moderate or more aortic regurgitation, significant mitral disease, those > 85 years of age, and those with significant coronary artery disease. Each attending physician selected surgery or close observation. Surgery was performed in 102, and the conventional approach was used in 95. The primary endpoint was operative mortality or cardiac death during a 500-day medical follow up. There were no significant differences between the two groups in age, sex, Euroscore, or ejection fraction at baseline. The primary endpoint was achieved in 18 of the conventional group and none of the operated group (24% vs. 0%, p < 0.001). In 57 propensity score-matched patients, the risk of all-cause mortality was lower in the operated group (HR = 0.135, 95% CI = 0.03-0.6, p = 0.008). In the conservative group, only aortic jet velocity ≥ 5 m/s was independently associated with cardiac death when adjusted for baseline Euroscore (HR = 4.76, 1.74-12.94, p = 0.002). The rate of progression of aortic velocity was the only independent predictor of cardiac mortality or surgery in the conservative group (HR = 9.75, 2.24-42.39, p = 0.002). The authors conclude that valve surgery in asymptomatic patients with very severe aortic stenosis is associated with improved survival compared to conventional therapy.


Most of us get very concerned when aortic valve area is < 0.80 cm2 in an asymptomatic patient. We might consider a careful exercise test to try to elicit symptoms, but the safety of this approach is not defined for very severe aortic stenosis. We might assess the severity of valve calcification, but it has not always been independently associated with mortality. Certainly we assess left ventricular function, but usually it is normal. So we follow the patient more closely and worry. This study suggests that early operation in asymptomatic patients with very severe aortic stenosis is associated with lower mortality than conventional management. However, there are several caveats to this conclusion that must be considered if one is going to heed this advice. First, the determination of the severity of aortic stenosis required a marked increase in aortic velocity (gradient) by echo Doppler in addition to a severely reduced valve area. This is a wise proviso because aortic valve area is markedly influenced by the left ventricular outflow tract area measurement, which is fraught with error. Second, the selected patients were low risk for surgery. The average age was 63 years. Their average ejection fraction was 65%, and their average Euroscore was about four. In addition, patients with ejection fraction < 50%, and those with significant coronary artery disease, were excluded.

The major limitation of this study was that it was a non-randomized, prospective, observational study and there were significant differences between the two arms. The surgical group had higher peak velocities and mean gradients, and there was a trend toward more bicuspid valves and less degenerative valves (p = 0.08). However, a propensity score-matched group showed similar results. Also, in more than six years of follow-up, there were no cardiac deaths in the surgical group, including zero operative mortality. So this was undoubtedly a skilled group of physicians and surgeons from two high-volume centers in Seoul, Korea.

The survival free of cardiac surgery in the conservatively managed group was 71% at two years and fell to 28% at six years. Their annual mortality was 1.7%/year. The unadjusted data showed that the rate of progression of aortic velocity and mean gradient, and aortic valve area, were significantly associated with cardiac event-free survival. Moderate-to-severe calcification exhibited a trend toward significance (p = 0.08). However, when adjusted for Euroscore, only the rate of progression of aortic peak velocity was significant. Thus, one could argue that aortic peak velocity is the key echo Doppler parameter to follow. This notion would be supported by other studies. In this study, velocities > 5 m/s were predictive; six-year survival was 85% when velocity was < 5 m/s and 62% when > 5 cm2. Other studies in broader patient populations have observed increased events when cutoffs of 3-4 cm2 are used. A 5 cm2 cutoff in asymptomatic patients seems reasonable if other criteria for surgery are met, such as low operative risk and a valve area estimation ≤ 0.75 cm2.