The Natural History of Asymptomatic Aortic Stenosis

ABSTRACT & COMMENTARY
A large cohort of patients with asymptomatic aortic stenosis, followed with clinical, exercise, and echo-Doppler data, has been reported and provides important information for the clinician. One hundred twenty-three adults with valvular aortic stenosis were enrolled between 1989 and 1995 and evaluated yearly over a mean of 2.5 ± 1.4 years. Nine died or underwent aortic valve replacement within the first study year, and 56 patients died or underwent aortic valve surgery at study completion. There were eight deaths; four were cardiac—all secondary to congestive heart failure. Careful clinical and functional assessments were carried out yearly. Fifty-two patients underwent coronary angiography during the study period for clinical reasons; half had coronary artery disease (CAD). Doppler-echocardiography was done yearly; imaging included calculations of LV volume and mass. Almost all of the patients underwent yearly maximal Bruce treadmill testing with Doppler-echo assessment of the aortic valve at baseline and immediately after exercise. The primary end point was death or aortic valve surgery. Assessment also included the rate of hemodynamic progression by echo, clinical status, and exercise test data.

Baseline characteristics: Seventy percent were men, with a mean age of 63 ± 16 years. Seventy percent had calcific degenerative aortic valve disease with co-existing mitral annular calcification in half. Left ventricular function was generally excellent. Mild aortic regurgitation (AR) was present in 75%, moderate-to-severe AR in 5%. Hemodynamic progression: Throughout the study period, aortic jet velocity increased by 0.7 m/s, with the mean aortic gradient increasing by 14 ± 13 mmHg; the aortic valve area decreased by -0.25 ± 0.28 cm2. The annual rate of change was 0.32 ± 0.34 m/s per year for aortic jet velocity, 7 ± 7 mmHg for mean gradient, and -0.12 ± 0.19 cm2 for aortic valve area. There was marked individual variability in the rate of hemodynamic progression. The rate of change in the first year was no different than the overall progression rate. Exercise testing was carried out with few complications, most of which were a fall in blood pressure. Seventy percent of subjects had abnormal S-T segments during exercise (not predictive of CAD), although half had an abnormal resting ECG. Individuals unable to exercise were older, had a lower ejection fraction, and had a substantially worse outcome. Clinical outcome was not benign. Of the initial 123 subjects, 7% died and 40% underwent aortic valve replacement. Primary indication for valve surgery was a decrease in exercise tolerance (40%), heart failure (23%), angina (15%), syncope (12%), and 6% were operated on for severe asymptomatic aortic stenosis. Kaplan-Meier event-free survival was excellent at one year at 93% ± 5% but fell dramatically to 26% ± 10% at five years.

Predictors of surgery or death were few and primarily related to aortic valve hemodynamic parameters. Furthermore, a more rapid rate of change in these measurements was related to clinical outcome. However, there was substantial overlap in individual values for all aortic valve parameters. Exercise performance was not predictive of outcome; while there were significant group differences between those with events and those without for a variety of exercise parameters, there was considerable overlap between groups and individual patients. Multivariate analysis demonstrated that aortic jet velocity at baseline, functional status at baseline, and the rate of change over time of jet velocity were predictors of clinical outcome. Of the group that died or underwent aortic valve replacement, the baseline aortic valve area was approximately 1 cm2 and the mean aortic jet velocity 4 m/s, both more severe than those without events. Patients with a baseline velocity of more than 4.0 m/s had a particularly adverse outcome, with the likelihood of remaining alive and without surgery at two years of only 20%. On the other hand, with an initial velocity of less than 3.0 m/s, the two-year event-free survival was 85%. A baseline velocity of 3.0-4.0 m/s had an intermediate event-free survival of 66% ± 13%.

Otto and colleagues conclude that the strongest predictor of clinical outcome was stenosis severity at baseline, which can be assessed by several echo-Doppler parameters. Functional status and the rate of hemodynamic progression were also strong predictors of outcome. Left ventricular mass and size, and echo diastolic parameters did not change appreciably. No clinical factors predicted progression. Otto et al stress the "very high rate of clinical events" in the overall group with initially asymptomatic aortic stenosis. By five years, 74% had died or undergone valve replacement, whereas after the first year, only 7% had an event. There were no sudden deaths, and the only cardiac deaths were related to severe LV dysfunction in the presence of CAD or refusal to undergo surgery in the presence of severe stenosis. Careful stress testing can be useful (and safe) in identifying individuals who will do well or who develop symptoms and impaired aerobic performance; failure to increase systolic blood pressure connotes an adverse outcome. Exercise function at entry was a predictor of outcome, but there was considerable overlap. Otto et al recommend that the measurement of aortic jet velocity be done annually, but meticulous echo-Doppler technique is required. In summary, asymptomatic individuals with an initial jet velocity of less than 3.0 m/s are unlikely to develop symptoms over the next five years, whereas those with a baseline velocity of more than 4.0 m/s have a less than 50% likelihood of death or aortic valve replacement within only 24 months. (Otto CM, et al. Circulation. 1997;95:2262-2270.)

COMMENT BY JONATHAN ABRAMS, MD

This is an outstanding example of careful clinical investigation. The results should benefit physicians who deal with patients with aortic stenosis. While it is somewhat surprising that the progression to surgery occurs so rapidly in asymptomatic patients with a baseline peak aortic valve gradient of more than 60 mmHg, other recent observations have indicated that occasionally similar subjects can progress rapidly to death or need for valve replacement. The study by Otto et al provides considerable assistance in how to follow such individuals. In an asymptomatic cohort of aortic stenosis, left ventricular function is usually quite good, as is the case in this study. Moderate aortic stenosis (aortic valve area of approximately 1 cm2), is a marker for more rapid progression over the next several years and should mandate a careful serial echo-Doppler approach for such individuals. Parameters of aortic stenosis severity, whether maximal velocity or aortic valve area, generally positively correlate with progression rates. Exercise testing, when carefully employed, can be useful in making a decision to hold off on surgery if the patient has good functional performance and no symptoms or may help select those individuals who should have surgical intervention earlier.

Physicians should choose a single echo lab that delivers high quality studies for the serial evaluations, and studies should be done on a yearly basis and carefully compared to previous studies. Progression of aortic stenosis equivalent to or greater than the median rate of progression in the Seattle cohort should prompt consideration of aortic valve replacement. Conversely, individuals who progress more slowly, particularly if exercise testing shows good performance, can continue to be observed; S-T segment analysis is of no practical use. It is reassuring that there were no sudden deaths in this group. In fact, onset of heart failure was associated with all cardiac deaths. Nevertheless, it is important to be aware that most individuals with moderate aortic stenosis who are asymptomatic will ultimately require surgical intervention. Aortic stenosis is clearly a dynamic process, and Otto et al list some of the current research indicating that inflammation, shear forces, lipid accumulation, and accumulation of T cells and fibroblasts may all relate to progression of aortic stenosis. In an accompanying editorial, Carabello provides a cogent overview of an approach to follow-up on patients with aortic valve stenosis, which should be recommended reading for all physicians who care for these individuals.