SAN FRANCISCO — After 7 years, transcatheter aortic valve replacement (TAVR) remains equivalent to surgery with respect to durability and safety in low-risk patients, a finding that may blunt surgery’s potential long-term advantage.

These results come from the latest analysis of the randomized PARTNER 3 trial, which compared TAVR with surgical aortic valve replacement in patients with aortic stenosis who have low surgical risk. At 1 year, TAVR was associated with superior outcomes, but 5-year data from the trial, which has a planned 10-year follow-up, showed an attenuation of this advantage. However, this did not prove an early signal that surgery is more durable, according to Michael J. Mack, MD, medical director for cardiothoracic surgery at Baylor Scott & White Health, Plano, Texas. 

“Valve function and durability over the 7 years were excellent and similar in both groups,” he reported.

TAVR No Longer Reserved for Nonsurgical Candidates

Once reserved for poor surgical candidates, TAVR is an attractive option for patients who wish to avoid surgery, Mack explained. In healthier patients with a longer life expectancy, the long-term results are critical to show that prosthetic valves delivered by a transcatheter approach offer durability comparable to surgery.

Mack presented the 7-year results of PARTNER 3 in a late-breaking clinical science session at the Transcatheter Cardiovascular Therapeutics (TCT) 2025 meeting. The findings were published simultaneously in The New England Journal of Medicine. 

In PARTNER 3, 1000 patients with severe symptomatic aortic stenosis but low surgical risk as assessed by the heart team at 71 participating centers were randomly assigned to TAVR with the SAPIEN 3 prosthetic valve (Edwards Lifesciences) or surgery with a bioprosthetic valve. After 1 year of follow-up, TAVR was associated with a 46% reduction in risk for the primary composite endpoint of death, stroke, or rehospitalization (hazard ratio [HR], 0.54; P = .001).

At 7 years, researchers assessed two primary endpoints. The first was the same nonhierarchical composite endpoint evaluated at 1 year. The second was a hierarchical evaluation of each endpoint using a win ratio methodology.

For the first primary endpoint, the rate favored TAVR over surgery (34.6% vs 37.2%; HR, 0.87; 95% CI, 0.7-1.08), though the nonsignificant 2.6% absolute difference was even narrower than the 4.6% absolute difference — also nonsignificant — observed at 5 years.

For the second primary endpoint, researchers calculated a win ratio for each outcome using 225,184 patient pairs. In this analysis, TAVR was linked to higher rates of death (19.5% vs 16.8%) and disabling stroke (5.1% vs 3.6%). However, it was also associated with lower rates of nondisabling stroke (3.7% vs 4.6%) and rehospitalization (20.6% vs 23.5%). 

When totaled, TAVR had wins in 28.8% of pairs, surgery in 27.6%, and the remaining 43.6% were ties. This produced a nonsignificant 1.04 ratio of wins in favor of TAVR (95% CI, 0.84-1.30; P = .7).

Equivalence Suggested Across Most Endpoints

For the 471 patients (95.0%) and 426 patients (93.8%) still available for follow-up at 7 years, equivalence of the aortic valve interventions was a recurring theme. 

For death, rates were higher with TAVR, whether the causes were cardiovascular (10.3% vs 7.8%) or noncardiovascular (10.2% vs 7.7%), but no clear patterns emerged in either category because of the small numbers. For example, sudden cardiac death was higher for TAVR (4 vs 1 patient), but death from myocardial infarction was lower (0 vs 2 patients).

Over 7 years, the rate of aortic reintervention was similar for the TAVR and surgery arms (6.7% vs 6%; P = .72), as were the rates of revascularizations (7.3% vs 7.7%; P = .57), any myocardial infarction (6% vs 5.6%; P = .96), and endocarditis (2.9% vs 2.8%; P = .89). However, clinical valve thrombosis was more common in the TAVR arm (2.8% vs 0.5%; P < .01), while new-onset atrial fibrillation was more common in the surgery arm (17.7% vs 43.5%; P < .0001).

The higher rate of new pacemaker implantation (17.3% vs 12.8%; P = .07) and the lower rate of serious bleeding (15.6% vs 18.5%; P = .14) with TAVR vs surgery approached statistical significance.

At 7 years, the mean gradient was 13.1 mm Hg after TAVR compared with 12.1 mm Hg after surgery (P = .02), and the aortic valve area was similar between treatment arms (1.93 cm² vs 1.84 cm²; P = .42). 

Although more TAVR patients experienced mild (28.7% vs 2.8%) and moderate or greater (0.8% vs 0%) paravalvular regurgitation at 30 days, the differences had diminished at 7 years. However, both mild (16.7% vs 1.6%) and moderate or greater regurgitation (1.0% vs 0.4%) remained more common in the TAVR arm.

As previously reported, the 1-year data from PARTNER showed that the length of hospital stay and recovery in general were faster after TAVR than after surgery. Quality of life scores measured by the Kansas City Cardiomyopathy Questionnaire (KCCQ) were also nonsignificantly higher at 1 year (90.7 vs 89.1). Over 7 years, Mack reported that KCCQ scores have remained more or less superimposable, with a slight, nonsignificant advantage for surgery at the most recent follow-up (86.2 vs 84.9). 

At 7 years, the proportion of patients who were alive at follow-up and had KCCQ scores greater than 75 was higher with surgery than with TAVR (63.3% vs 59%; P = .26), as was the proportion of surgery patients alive with a durable valve (75% vs 73.4%; P = .63). Neither result was statistically significant.

“What we are seeing in these results are very similar outcomes across a variety of methodologies to evaluate these results,” said Stefan Blankenberg, MD, cardiovascular interventionalist and professor of medicine at the University Heart Center in Hamburg, Germany.

A Potential Win for TAVR?

Like others, Blankenberg, who served as a discussant for the PARTNER 3 trial, suggested this is reassuring for either choice of valve treatment, but it will be considered a potential win for patients hoping to avoid surgery.

However, even after 7 years, the surgical approach still has the advantage of longer follow-up. Mack said valve survival of at least 10 years after surgery is common. The same type of follow-up is not yet available for TAVR. 

For younger patients with the greatest likelihood of long-term survival, 10-year results will be important if they tell a different story, but Michael J. Reardon, MD, chief of cardiothoracic surgery at Baylor College of Medicine in Houston, said he was impressed with the lack of a difference in failure rates in the follow-up so far.

He and others expect the 7-year PARTNER 3 results to be helpful when discussing the merits of TAVR relative to surgery in patients with a failing aortic valve.

PARTNER 3 received financial support from Edwards Lifesciences. Mack reported having financial relationships with Abbott, Edwards Lifesciences, and Medtronic. Blankenberg reported having financial relationships with Bristol Myers Squibb, Boehringer Ingelheim, Daiichi-Sankyo, and Eli Lilly. Reardon reported having financial relationships with Boston Scientific, High Life, JValve, Medtronic, Siemens, and W.L. Gore and Associates.