Early TAVR cut heart failure events in moderate aortic stenosis

Early TAVR cut the rate of death or a heart failure event at two years from 23.9% to 16.7% in people with moderate aortic stenosis and at-risk features, in the PROGRESS trial. That is a 33% relative reduction and a 7.2 percentage-point absolute gap (95% CI -13.1 to -1.3; p=0.02 for superiority). Heart failure hospitalisations roughly halved. All-cause death was similar between arms.

The late-breaking results were presented on 2 November 2026 at Transcatheter Cardiovascular Therapeutics (TCT) 2026 in San Diego and published the same day in the New England Journal of Medicine. Edwards Lifesciences, which funded the trial and makes the SAPIEN 3 platform used in it, issued a matching press release. Use of SAPIEN 3 systems for moderate aortic stenosis remains investigational; safety and effectiveness for that use have not been established for labelling.

What moderate aortic stenosis is

Aortic stenosis is narrowing of the aortic valve, which restricts blood leaving the left ventricle. Guidelines have long treated moderate disease as milder than severe disease and managed many people with clinical surveillance: check-ups and echocardiography every one to two years, with valve replacement deferred until the valve becomes severe or clear guideline triggers appear.

Moderate disease is not one tidy group. About half of people with moderate aortic stenosis are estimated to have at-risk features such as symptoms, cardiac damage or dysfunction. Those features track with poorer outcomes or faster progression. PROGRESS asked whether proactive TAVR helps that subgroup sooner.

How PROGRESS was built

PROGRESS (NCT04889872) is a prospective, open-label, multicentre randomised trial. Sites enrolled 744 people aged 65 or older with moderate calcific aortic stenosis confirmed by an independent echocardiography core laboratory, plus at least one at-risk feature. Randomisation was 1:1 to transfemoral TAVR with a valve from the Edwards SAPIEN 3 platform (SAPIEN 3, SAPIEN 3 Ultra or SAPIEN 3 Ultra RESILIA) or guideline-recommended clinical surveillance with deferred aortic valve replacement when indicated.

Enrollment ran from October 2021 to January 2024 at 74 centres across six countries, including the US, Japan, Australia, Europe and Canada. Of 1,294 people screened, 550 (42.5%) were excluded, most often because they did not have moderate aortic stenosis or had anatomy unsuitable for TAVI. More than 90% of those randomised had concordant moderate aortic stenosis on guideline echo criteria.

At-risk features in the protocol included valve-related symptoms; left ventricular ejection fraction under 60%; diastolic dysfunction grade 2 or higher; stroke volume index under 35 ml/m2; atrial fibrillation in the prior six months; NT-proBNP more than three times the upper limit of normal; or a high aortic valve calcium score on CT. Most participants (72%) had two or more of these features; 40% had three or more.

Who took part

Mean age was about 78 years. Women made up 39.1% of the cohort. Mean aortic valve area was 1.2 cm2 and mean peak velocity 3.3 m per second. Mean STS score was about 3%, and a heart team judged 45.6% to be at low surgical risk. More than 95% were symptomatic at baseline; more than one third had NYHA class III or IV symptoms. More than 90% had preserved left ventricular ejection fraction (50% or higher). Mean baseline Kansas City Cardiomyopathy Questionnaire (KCCQ) overall summary score was 64.2.

The most common at-risk feature was NYHA class II or worse symptoms (96.4%), followed by diastolic dysfunction (43.8%), a high aortic valve calcium score (25.7%) and LVEF under 60% (25.0%).

The primary result

The primary effectiveness end point was a composite of death from any cause or a heart failure event with or without hospitalisation at two years, tested for superiority in the intention-to-treat population.

ACC coverage of the presentation puts the risk difference at -7.2 percentage points (95% CI -13.1 to -1.3) with p=0.02 for superiority. The benefit on the primary end point was consistent across pre-specified subgroups. A hierarchical secondary composite of death, stroke or heart failure hospitalisation at two years occurred in 17.7% on TAVR and 23.3% on surveillance.

Put another way: for every roughly 14 people treated with early TAVR instead of surveillance, one fewer primary event occurred over two years (absolute risk reduction 7.2 percentage points). That arithmetic uses the Kaplan-Meier estimates reported with the trial; it is a rough number of needed to treat, not a formal NNT from the paper.

Quality of life and delayed valves

TAVR patients gained more on quality of life. A higher share were alive with at least a 10-point improvement in KCCQ overall summary score sustained through one and two years: 39.0% versus 23.2% (p<0.001). Edwards also reported a 15.5-point KCCQ improvement at 30 days after TAVR.

Surveillance was not static. In that arm, 27% underwent delayed aortic valve replacement within one year and 57.4% within two years (204 people). About 80% of those delayed procedures were done after progression to severe aortic stenosis, and most (81.4%) met guideline class I indications. Before delayed replacement, health status fell; scores rose again by 30 days after the later procedure. That pattern is one reason investigators argue waiting is not free of cost for this selected group.

Safety at 30 days

The primary safety end point was a 30-day composite of death, stroke, life-threatening or fatal bleeding, stage 4 acute kidney injury, hospitalisation due to a device- or procedure-related complication, or valve dysfunction needing reintervention. In the as-treated TAVR group the rate was 5.9%, below the performance goal of 13.7% (p<0.001). The most common component was hospitalisation related to complications (5.6%). Investigators highlighted 0% mortality and 0% disabling stroke at 30 days among TAVR-treated patients in this moderate-stenosis cohort.

Procedure-related adverse events did not differ meaningfully between people who received TAVR up front and people in the surveillance arm who later underwent aortic valve replacement, according to the NEJM abstract summary.

How this sits next to earlier TAVR timing trials

PROGRESS extends a line of work on intervening before classic "severe, symptomatic" thresholds. The EARLY TAVR trial, also led by Philippe Genereux and funded by Edwards, previously showed that early TAVR beat surveillance for asymptomatic severe aortic stenosis on a composite of death, stroke or unplanned cardiovascular hospitalisation. PROGRESS moves one step earlier on the stenosis scale, into moderate disease, but only in people who already have symptoms or structural or functional risk markers.

Lead investigator Philippe Genereux of Atlantic Health Morristown Medical Center told TCTMD that moderate aortic stenosis "is not benign" and that these patients should be sent to a valve clinic for a full assessment. Discussant Stephan Windecker of Bern University Hospital argued that clinicians should move past a binary severe versus non-severe label and weigh risk along a continuum.

Catherine Otto of the University of Washington, in an accompanying NEJM editorial summarised by TCTMD and ACC, cautioned that the enrolled patients were atypical because moderate aortic stenosis does not usually cause symptoms or left ventricular dysfunction. She wrote that TAVR is not appropriate for most adults with moderate aortic stenosis, though it may be considered in selected older patients at high risk of adverse outcomes when other approaches are not effective. Confirmatory studies, longer follow-up, durability data and cost analyses are still needed.

Surgeon Tsuyoshi Kaneko of Washington University in St. Louis noted that people with anatomy poorly suited to TAVR (for example heavily calcified bicuspid valves) were excluded, and that surgical valve replacement may be the better tool for some low-comorbidity patients with unsuitable TAVR anatomy.

Limits to keep in view

• Industry-funded, open-label design. Edwards funded PROGRESS and manufactures the valves studied. Symptom and quality-of-life end points can be influenced when participants know their assignment.
• Selected population. Almost everyone was symptomatic. Results do not apply to quiet, low-risk moderate stenosis without at-risk features.
• Anatomy filter. Only TAVR-suitable anatomy entered. Screen failures for unsuitable anatomy were common.
• Mortality unchanged at two years. The primary win was driven by fewer heart failure events and hospitalisations, not by fewer deaths.
• Two-year window. Participants will be followed for up to 10 years. Durability, late valve failure and longer survival curves are still ahead.
• Regulatory status. Moderate aortic stenosis is an investigational indication for SAPIEN 3 systems. Label expansion would need regulators to review the full evidence package.
• Editorial caution. Otto's accompanying comment stresses that most people with moderate aortic stenosis should not be routed to TAVR on the basis of this trial alone.

What readers and clinicians can do with this now

People told they have moderate aortic stenosis, especially with breathlessness, reduced exercise capacity, raised biomarkers, atrial fibrillation or imaging signs of heart strain, can ask for referral to a multidisciplinary valve clinic instead of defaulting to a one-to-two-year echo gap. A second look sometimes reclassifies severity. Shared decisions still weigh age, anatomy, surgical risk, life expectancy and patient goals.

For context on how trial composites and absolute risks are read, see our guide to absolute risk in clinical trial results. Related valve and heart-failure coverage on this site includes tricuspid repair in heart failure (TRIC-I-HF) and aficamten in nonobstructive hypertrophic cardiomyopathy. For anticoagulation decisions that often sit alongside structural heart care, see SINGLE AF at intermediate stroke risk.

This article explains published research and company announcements. It is not medical advice. Talk to your own doctor or heart team before changing any treatment.