The Natural History of Aortic Stenosis Revisited

By Michael H. Crawford, MD, Editor

SYNOPSIS: A large multi-institution observational study of patients referred for Doppler echocardiography to assess for aortic stenosis has shown that discrepant measurements are not uncommon. When four-year all-cause untreated mortality is considered, the intermediate grades of aortic stenosis behave like the next highest level stenosis, which suggests that we should consider intervening earlier in moderate to severe stenosis.

SOURCE: Généreux P, Sharma RP, Cubeddu RJ, et al. The mortality burden of untreated aortic stenosis. J Am Coll Cardiol 2023; Oct 20. doi: 10.1016/j.jacc.2023.09.796. [Online ahead of print].

Severe aortic stenosis (AS) with either symptoms or reduced left ventricular (LV) function is the only class one guideline criterion for aortic valve replacement (AVR). The determination of the severity of AS by Doppler echocardiography often is challenging, yet underestimation of severity may lead to poor outcomes. Thus, this study of the natural history of AS of all severities is of interest.

From a 24-institution prospectively collected database, 1.7 million echo reports from 1.1 million patients were subjected to a natural language processing algorithm to detect studies where AS was evaluated between 2016 and 2022. After excluding those with missing data, 595,120 patients were categorized for AS severity using standard criteria. The primary outcome was all cause untreated mortality, censored for death, the last clinical encounter, and AVR. The secondary endpoint was time to AVR. Most patients evaluated by echo did not have AS (88%). The remaining 70,778 patients (12%) had the diagnosis of AS categorized as mild in 6%, mild to moderate in 1.0%, moderate in 2.4%, moderate to severe in 0.6%, and severe in 2%. Patients categorized as having AS and having more severe AS had more comorbidities, such as hypertension, coronary artery disease, atrial fibrillation, low LV function, and concomitant disease of other valves.

After a median follow-up of four years, mortality occurred in 14% of those with no AS, 25% of those with mild AS, and progressed to about 45% in those with moderate to severe or severe AS. Treatment rates were 1% in mild AS, 4% in mild to moderate AS, 11% in moderate AS, 37% in mild to moderate to severe AS, and 61% in severe AS. An adjusted multivariate analysis showed that all severities of AS were associated with increased mortality if untreated. The authors concluded that mortality is high across all severities of AS, and AVR rates are low for patients with severe AS.

COMMENTARY

The American College of Cardiology/American Heart Association guidelines on valvular heart disease categorize AS as mild, moderate, or severe. Often, the three critical Doppler echo parameters that are used to categorize the severity of AS are discordant: peak aortic valve (AV) velocity, mean pressure gradient, and calculated aortic valve area. When this cannot be technically resolved, intermediate severity grades are chosen (mild to moderate, moderate to severe). Moderate to severe AS usually is not considered severe enough to recommend valve replacement unless other compelling factors are present, such as a dobutamine echocardiogram study that shows severe AS with increased flow, elevated biomarkers (brain natriuretic peptide, troponin), marked myocardial fibrosis on cardiac magnetic resonance imaging, or a severely calcified valve on a computed tomography scan.

The Généreux et al study is the largest study ever published and spans the whole range of AS severity in this echo for possible AS population. It clearly shows that intermediate severity of AS is not uncommon, with 8% categorized as mild to moderate and 5% categorized as moderate to severe. Perhaps most interesting is that these intermediate categories had almost the same mortality without AVR over four years as the next highest grade of AS. In particular, the mortality of mild to moderate AS was 30% and that of moderate AS was 34%. Likewise, the mortality of moderate to severe AS was 46% and that of severe AS was 45%. Thus, the authors postulated that we should treat the intermediate grades of AS as if they are the next highest grade, which would mean considering AVR in patients with moderate to severe AS by Doppler echocardiogram. Of course, the basic premise of any therapy is that the risk of the therapy be lower than the risk of the natural history of the disease. In the last century, when treatment for AS meant surgical AVR, that might have excluded considering AVR for moderate to severe AS, but now, with transcatheter (T) AVR established, it makes more sense.

In the Généruex database, the rate of AVR was low in moderate AS (11%) and severe AS (61%). Given our current paradigm, the moderate AS AVR rate would be expected, but the severe AS AVR rate seems lower than expected. However, some of the severe AS patients may have been asymptomatic or had normal LV function. Some may have refused treatment, and in some treatment, may have been considered futile because of the condition of the patient (demented, frail). These data are not provided in the Généruex study.

Another weakness of this study is that the cause of death is not provided. Given the frequency of comorbid conditions, it is possible that much of the mortality observed was caused by other conditions rather than AS per se. Also, they used a natural language algorithm to extract the severity of AS from the echocardiographic reports rather than analyzing the actual images. In addition, the population is biased in that only patients referred for echocardiogram to evaluate for AS were included. There may have been technical variance between the different institutions involved and there was no more granular analysis of the echocardiograms, such as stroke volume index.

Finally, medical records may not capture all the deaths in any population. Although imperfect, the study does add support to the notion that, in the TAVR era, earlier intervention in AS may be beneficial. Better yet, we need to find a way to prevent the development of AS.