Despite Positive Trials, Early Surgery in Asymptomatic AS Should Wait

John Mandrola, MD


November 15, 2021

Now there are not one but two 'positive' randomized controlled trials on early surgical aortic valve replacement in patients with severe but asymptomatic aortic stenosis (AS).

You may think that is enough to change the practice of watchful waiting in these patients, but I will argue that it is not.

At the American Heart Association meeting, Marko Banovic, MD, PhD, from the University of Belgrade in Serbia, presented results of the AVATAR trial comparing watchful waiting or surgery in 157 patients with severe but asymptomatic AS. The journal Circulation simultaneously published the study results.

South Korean investigators first challenged the dogma of watchful waiting with publication of the RECOVERY trial, which found a 91% reduction in operative death or death due to cardiovascular causes during the 6-year follow-up. The limitations of RECOVERY included its small numbers of patients and the question of whether these patients were truly asymptomatic.

AVATAR investigators added an important twist: To be enrolled, patients had to achieve 85% of their age-predicted maximum heart rate on exercise testing — which ensured a truly asymptomatic cohort.

The Data on Early Surgery in AS

Patients were young (average age 67 years) and had low surgical risk (median STS-PROM score 1.7%). The cause of AS was degenerative in 85% of cases and bicuspid in 14%.

Perhaps the most important characteristic was that these patients were well enough to meet the exercise testing entry criteria.

Over a median follow-up of 32 months, the primary endpoint of all-cause death, myocardial infarction, stroke, or heart failure hospitalization occurred in 13 of 78 (16.6%) patients in the early surgery arm vs 26 of 79 (32.9%) in the conservative arm.

This translated to a hazard ratio of 0.46 (95% CI, 0.23 - 0.90; P =.02). The Kaplan-Meier curves for the primary endpoint separated at about a year.

The composite endpoint was driven by all-cause death; 16 deaths in the conservative arm vs 9 in the early surgery group. But cardiovascular death rates were not different (approximately 9% in both arms).


AVATAR favored the early surgery approach. The driver of the composite endpoint was death — an unbiased endpoint. And the highly positive results from RECOVERY should bolster confidence in AVATAR. The easy take-home message would be that dogma should change and patients with severe AS should be offered surgery before symptoms.

But I side with the AVATAR authors who wrote this about their results: "the trial findings will require further confirmation in a larger study."

The two reasons I agree with the authors are fragility and external validity

Fragility of the AVATAR Results

AVATAR results are fragile. If only two patients in the active arm converted from not having the primary endpoint to having the primary endpoint, the study would lose statistical significance.

The fragility becomes apparent when comparing cause of death in the two groups. Since these patients were young, had low surgical risk, and were well enough to pass an exercise test, you can safely assume they had low comorbidity and hence few competing causes of death.

You would expect a reduction in cardiovascular deaths to drive the difference in mortality, but that is not what AVATAR found.

In the supplement, the authors list the causes of death. There were three COVID-19 pneumonia deaths in the conservative arm and none in the early surgery arm. Although there remains much to learn about SARS-CoV-2, it's a stretch to posit that aortic surgery would prevent death from a viral infection. This was likely a chance finding.

There was also a small number of total primary outcome events.

In addition to a difference of only 7 deaths, there were just 6 strokes and myocardial infarctions combined in both arms. Heart failure hospitalizations favored the conservative arm (7 vs 1), but this is not a great endpoint because the whole point of the conservative approach is to wait for symptoms — of which heart failure is one.

Small event numbers was also the main problem in RECOVERY. The difference in cardiovascular deaths between groups in that trial was 10; and an inspection of these deaths also suggests fragility. Three of the 11 total CV deaths in the watchful waiting arm were unrelated to a conservative approach; two were due to complications from redo aortic valve replacements, presumably crossovers, and one death was due to an intracranial bleed.

The fact that just a handful of CV death adjudications could flip a trial from neutral to reaching statistical significance means that we should be careful using these data to change established practice.

External Validity

The other reason we need to get more data is external validity or generalizability. The average age of patients in AVATAR was 68 years and more than 80% had degenerative disease. But aortic stenosis is mostly a disease of the elderly, who often present with multimorbidity.

The typical patient I see with nonbicuspid AS is more likely an 80-year-old patient with poor mobility due to arthritis, chronic kidney disease, previous bypass surgery, and maybe a pacemaker.

Older age and comorbid conditions could affect both the efficacy and risk of an intervention such as open-heart surgery. Although transcatheter aortic replacements have been found noninferior to surgical aortic valve replacement, the trials were done in symptomatic patients and therefore may not apply to patients without symptoms.


Given our oath to first, do no harm, preventive procedures used in people without complaints should pass a high bar of evidence.

When the preventive procedure involves a sternotomy and cardiopulmonary bypass the bar must be super high.

Despite these two positive trials, we still need more data before upending the conservative approach in most patients with AS and no complaints. 

John Mandrola, MD, practices cardiac electrophysiology in Louisville, Kentucky and is a writer and podcaster for Medscape. He espouses a conservative approach to medical practice. He participates in clinical research and writes often about the state of medical evidence. 

Follow John Mandrola on Twitter

Follow | Medscape Cardiology on Twitter


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.