Asymptomatic severe aortic stenosis: three patients, two decisions

What changed

The 2025 ESC/EACTS guideline gives early intervention a Class IIa, Level A recommendation in genuinely asymptomatic severe high-gradient aortic stenosis with ejection fraction of 50% or more and low procedural risk. It is offered as an alternative to close active surveillance, not a replacement for it. ^[1]

That sentence ends a long argument about whether we are allowed to treat these patients early. It opens a harder one. The expert question is no longer whether early valve replacement is permissible. It is which asymptomatic patient should be treated now, what kind of benefit they can expect, and what first intervention leaves them the best lifetime strategy.

Read EARLY TAVR properly

The largest trial behind that shift randomised 901 patients, with a mean age of 76 and mean STS risk of 1.8%, to early transfemoral TAVI or surveillance. If you remember four numbers from it, remember these: 901, 3.8 years, 26.8% versus 45.3%, and 87%. ^[2]

The relative hazard of death, stroke or unplanned cardiovascular hospitalisation was halved over a median of 3.8 years. Dissect the composite, though, and the picture changes. Death was 8.4% versus 9.2%. Stroke was 4.2% versus 6.7%. Unplanned cardiovascular hospitalisation was 20.9% versus 41.7%. The trial's benefit was driven mainly by preventing deterioration and admission, not by a demonstrated gain in survival.

Then the fourth number. By the end of follow-up, 87% of the surveillance arm had undergone valve replacement, and more than a quarter had crossed over within six months. EARLY TAVR did not compare treatment with no treatment. It compared early TAVI with a strategy that usually led to later valve replacement. At the trial horizon, surveillance mainly postponed the procedure rather than avoiding it.

So when the 78-year-old asks me what treating now will do for him, I say something like this:

“The strongest trial evidence says that treating now substantially reduces the chance of an unplanned cardiovascular admission or deterioration leading to later treatment. It has not shown, over the follow-up we have, that treating now helps you live longer.”

That distinction matters to patients. Some hear it and want the valve tomorrow. Others hear it and would rather wait while they feel well. Both are reasonable responses to the same honest sentence.

The surgical trials are a different population

The survival argument comes from surgery. RECOVERY, now reported at ten years, randomised 145 patients with very severe stenosis to early surgery or conservative care. Operative mortality or cardiovascular death occurred in 3% versus 24%, and all-cause death in 15% versus 32%. ^[3] AVATAR's long-term follow-up also showed lower all-cause mortality with early surgery. ^[6]

These are strong results. But do not merge these populations with EARLY TAVR. RECOVERY's patients were in their mid-sixties, with very severe disease, and a majority had bicuspid valves. Many of them would already meet the guideline's separate criteria for very severe stenosis. When all four randomised trials are pooled, early intervention reduces hospitalisation and stroke, but not all-cause or cardiovascular mortality. ^[5]

EVOLVED is the necessary counterweight, and it is our own. This UK and Australian trial selected patients with myocardial fibrosis on cardiac MRI and found no significant reduction in death or aortic stenosis-related hospitalisation with early intervention. ^[4] It is an argument against treating everyone early, and against imaging biomarker triggers used as simple rules.

The guideline's adverse features still matter: very high velocity, severe calcification with rapid progression, markedly raised natriuretic peptides, and ejection fraction below 55%. I treat them as features that tip the balance towards intervention, not as mandates that remove judgement. ^[1]

Two decisions, not one

Here is a proposition I hear increasingly often:

“EARLY TAVR means an asymptomatic 68-year-old with severe high-gradient tricuspid aortic stenosis and a negative exercise test should now have TAVI.”

I would reject it.

EARLY TAVR strengthens the case for treating that patient early. It does not establish whether TAVI is the right first valve for him, because it had no surgical comparator. It also tested one balloon-expandable platform and was funded by its manufacturer. The guideline itself preserves surgery as the preferred strategy for low-risk patients under 70 and recommends TAVI primarily from 70 with suitable anatomy. ^[1]

EARLY TAVR informs the first decision far more than it answers the second.

A structural colleague will reasonably ask why a healthy 68-year-old should face a sternotomy when TAVI outcomes are excellent and a second transcatheter valve can follow later. The weakness is the assumption that TAVI followed by TAVI-in-TAVI is a proven lifetime strategy. It is not. At 68, the unanswered questions include durability beyond current trial horizons and coronary re-access after one and then two transcatheter valves. They also include sinus sequestration, the feasibility of the second valve, and what happens if that strategy fails and explant becomes necessary.

A technically easy first intervention is not necessarily the best first move in a 20-year valve strategy.

For my 68-year-old, early valve replacement is now reasonable. So is close surveillance, because he has none of the adverse features. If he chooses to proceed, I want a surgery-first conversation, not a reflex TAVI. For my 78-year-old, who sits close to the EARLY TAVR population, I now regard early TAVI as a strong clinical option for Heart Team discussion rather than waiting for him to deteriorate. In England, that clinical conclusion does not by itself create a commissioning route.

The patient who cannot tell us

My 83-year-old is where the word “asymptomatic” does the most damage.

Calling her asymptomatic is intellectually unsafe. She has no breathlessness because she no longer asks her body for anything that would reveal it. EARLY TAVR enrolled a selected, low-risk population. Her competing illness, limited activity and probable frailty put her well outside its centre of gravity.

The question for her is not whether severe stenosis warrants TAVI. It is how much of her limitation and prognosis is caused by the valve, and how likely a technically successful procedure is to give her a life she would recognise as better. Frailty, cognition, renal trajectory, rehabilitation potential and her own goals matter more here than another decimal place of valve area.

A technically successful TAVI that leaves her alive but functionally unchanged may still represent procedural success and therapeutic failure.

Where England sits

The European guideline is the specialist authority for this decision, but English practice runs on other documents too, and they have not caught up.

NICE NG208 still sets a narrower threshold for referral for intervention in asymptomatic severe stenosis. It asks us to consider referral only with a velocity above 5 m/s, valve area below 0.6 cm², ejection fraction below 55%, natriuretic peptides above twice the upper limit of normal, or symptoms unmasked on exercise. ^[7] NICE last reviewed NG208 in October 2025 and confirmed that an update of its TAVI recommendations is in development. ^[8]

Commissioning is the sharper problem. NHS England's version 2 position statement, updated in May 2023, applies to symptomatic severe aortic stenosis. My 78-year-old, the patient the new evidence most clearly supports, falls outside it on symptom status. ^[9]

I do not think the answer is to wait quietly for him to become symptomatic so that he fits a policy. Nor is documentation alone enough. A divergence between evidence and commissioning is not a reason to hide the decision. It is a reason to make the governance explicit.

There is one more uncomfortable point. In a well-run valve clinic that genuinely reviews patients every six months and treats promptly when symptoms appear, the extra benefit of pre-emptive intervention is probably smaller than the trials suggest. Where follow-up slips or the list after a symptomatic trigger runs to months, it is probably larger. The guideline says as much, naming long waiting lists as an argument for early intervention. Our own services are part of the evidence. ^[1]

The UK is also generating its own answer. EASY-AS, a UK-based trial randomising asymptomatic severe aortic stenosis to early intervention or surveillance, is still running. ^[10] Until it reports, we are making this decision with good but incomplete evidence, and we should say so to patients.

What should change on Monday

Do not let “asymptomatic severe aortic stenosis” end the reasoning with “repeat echo in six to twelve months”.

1. 01 Verify symptoms properly, with an exercise test where the patient can do one.
2. 02 Look at progression, calcification, natriuretic peptides and the left ventricle's response.
3. 03 Weigh competing risk.
4. 04 Decide on candidacy for early valve replacement.
5. 05 Then, separately, decide the lifetime strategy.

The separation between the indication to intervene and the choice of first valve is the most important practical idea in this field right now.

Back to the clinic

My 68-year-old goes home with a real choice and a surgical opinion if he wants to proceed. My 78-year-old is offered a Heart Team discussion about early TAVI, with an honest account of what it does and does not buy and a clear explanation that the commissioning route must be resolved. My 83-year-old gets a frailty assessment and a conversation about what she wants her next few years to look like, before anyone measures her annulus.

Three patients who all said they felt well. Three different right answers.