New 10-Year TAVI Data Generally ‘Reassuring’ but Highlight Challenges

With an updated valve design, balloon-expandable TAVI appears to provide similar long-term results when compared with SAVR in patients with symptomatic severe aortic stenosis and intermediate surgical risk, according to 10-year follow-up of the PARTNER 2A trial and the PARTNER 2 Sapien 3 Intermediate Risk Registry (P2S3i).

In PARTNER 2A, rates of all-cause mortality and aortic valve reintervention at 10 years were higher in patients who underwent TAVI with the second-generation Sapien XT valve (Edwards Lifesciences) versus those treated surgically, researchers led by Vinod Thourani, MD (Piedmont Heart Institute, Atlanta, GA), report.

However, those outcome differences were not observed in a prespecified, propensity score-matched comparison of patients who underwent TAVI with the third-generation Sapien 3 device (Edwards Lifesciences) in the P2S3i study compared with the SAVR-treated patients from PARTNER 2A, suggesting that enhancements in valve design and use of preprocedural planning with CT improved results with TAVI, Tamim Nazif, MD (NewYork-Presbyterian/Columbia University Irving Medical Center, New York, NY), and colleagues report.

Both analyses were published in the June 16, 2026, issue of JACC.

Previously reported 5-year results of pivotal clinical trials of TAVI with different valve platforms across surgical risk levels have indicated that the procedure performs well versus SAVR in the short-to-intermediate term, Nazif told TCTMD.

“Taken together, the studies show a likely early advantage of TAVR with respect to the important primary clinical endpoints, with a later catch-up of surgery, such that there are no major differences between the therapies at 5 years,” he said. “However, we know that 6 to 7 years represents the beginning of a period of vulnerability for early structural valve deterioration of bioprosthetic valves.”

Therefore, 10 years is a key time point for the assessment of TAVI durability. Before these new analyses, extended follow-up to that point in a randomized comparison of TAVI versus SAVR had been reported only from the NOTION trial, which was limited by a relatively small sample size, inclusion of mostly lower-risk patients, and use of a first-generation self-expanding transcatheter valve.

“As TAVR continues to expand and as we begin to consider strategies for lifetime management of our younger and lower-risk patients, there remains a critical unmet need for more and better data regarding transcatheter heart valve durability, and I think that these current studies are an important step forward in that regard,” Nazif said, adding that the field is anxiously awaiting similar data on lower-risk patients from PARTNER 3 and the Evolut Low Risk Trial.

Michael Borger, MD, PhD (Leipzig Heart Center, Germany), who wrote an editorial accompanying the two studies, told TCTMD that with these new findings, “we’ll be able to make more valid comments about long-term durability, at least in this intermediate-risk patient group.”

He said: “This is data that we’ve all been looking forward to seeing since we know from the surgical literature that structural valve deterioration tends to increase exponentially between 5 and 10 years after implantation.” That depends on patient age, however, with individuals over 80 years old—like many of those in the current studies—expected to have negligible amounts of structural valve deterioration after SAVR.

Therefore, this new information is “extremely interesting, but really we need the 10-year results of the low-risk trials to fully evaluate the freedom from structural valve degeneration and durability of these TAVR prostheses, especially the more-recent generations,” Borger said.

Outcomes After a Decade

In PARTNER 2A, patients with an intermediate surgical risk were randomized to TAVI with the Sapien XT valve or SAVR, stratified by anatomical suitability for transfemoral or transthoracic (ie, transapical or transaortic) access. TAVI was shown to be noninferior to surgery at 2 years, with similar results seen at 5 years.

Although the trial was initially designed to follow patients for 5 years, the US Food and Drug Administration later mandated tracking patients for 10 years due to concerns about valve durability, and patients were asked to re-consent to additional follow-up. The researchers also performed a vital status sweep for patients who didn’t agree to additional follow-up, withdrew, or were lost to follow-up.

The 10-year analysis of PARTNER 2A included 1,910 patients (mean age 82 years; 45.4% women) who received a valve, including 974 who underwent TAVI and 936 who underwent SAVR. Mean STS score at baseline was 5.8%.

Before the vital status sweep, 10-year follow-up was available for 61.9% of patients in the TAVI group and 54.1% of those in the SAVR group. The vital status sweep allowed for an assessment of all-cause mortality in about 90% of patients.

Overall, the mortality rate was higher in the TAVI group (86.1% vs 82.8%; HR 1.13; 95% CI 1.02-1.25), driven by a difference in the 23.1% of patients who underwent the procedure via transthoracic access (93.2% vs 85.1%; P < 0.01). There was a similar rate of mortality in the TAVI and SAVR groups among patients suitable for transfemoral access (83.9% vs 82.1%; P = 0.27).

After accounting for the competing risk of death, site-reported aortic valve reinterventions also were more frequent with TAVI through 10 years (6.3% vs 1.6%; P < 0.001), with most of these interventions done in response to restenosis and aortic regurgitation (AR).

Echocardiographic information in a limited number of patients indicated that mean aortic valve gradients remained stable between 5 and 10 years of follow-up, with mean gradients of 12.6 and 12.7 mm Hg, respectively, reported in the TAVI and SAVR arms. Patients who had been treated with TAVI compared with surgery were more likely to have moderate or greater total AR (12.5% vs 2.7%) and moderate or greater paravalvular regurgitation (10.5% vs 0) at 10 years.

“These results should be interpreted in the context of challenges associated with long-term follow-up, including trial attrition, differential rates of follow-up, and the competing risk of mortality in an aging population,” Thourani et al conclude.

The P2S3i Registry

The P2S3i registry was created because the third-generation Sapien 3 valve was released shortly after PARTNER 2A began randomizing patients to TAVI with the second-generation Sapien XT device. The registry had the same eligibility criteria as the trial but also required CT imaging for anatomic assessment, valve sizing, and vascular access evaluation; there was a prespecified goal to compare outcomes with the SAVR arm of PARTNER 2A after propensity score matching. Registry patients also were asked to re-consent to extended follow-up out to 10 years.

After propensity score matching, there were 783 patients in the TAVI group and an equal number in the SAVR group (mean age 82 years; 43% women). Mean STS score at baseline was 5.5%.

Follow-up at 10 years was available for 59.8% of patients who underwent TAVI and 53.6% of those treated surgically, with vital status sweeps allowing for an assessment of all-cause mortality in nearly 90%.

In contrast to the findings of PARTNER 2A, there was no difference in the rate of all-cause mortality at 10 years between the TAVI and SAVR groups (83.4% vs 82.3%; HR 1.01; 95% CI 0.91-1.13), with similar results irrespective of the type of vascular access.

The aortic valve reintervention rate, adjusted for competing mortality risk, also did not significantly differ between the TAVI and SAVR groups (2.0% vs 1.9%; P = 0.47).

Echocardiography in a small number of patients showed similar mean gradients in the TAVI and surgery groups at 10 years (11.0 vs 12.6 mm Hg), with a higher proportion of TAVI-treated patients having mild AR (33.3% vs 16.1%) and mild paravalvular regurgitation (21.9% vs 3.4%), as seen at earlier time points. No moderate or greater regurgitation was reported.

Nazif said the better long-term TAVI results seen in the registry versus PARTNER 2A are likely related to improvements made to the Sapien 3 valve versus the earlier Sapien XT device—including a redesigned stent frame, changes to the leaflet configuration, an enhanced skirt to cut down on paravalvular regurgitation, and smaller catheters to facilitate transfemoral vascular access—and the addition of routine preprocedural planning with CT.

In short, TAVI as performed in the P2S3i registry more closely reflects contemporary practice, in which Sapien 3 has been phased out and operators are now using the balloon-expandable S3 Ultra and Ultra Resilia valves incorporating further improvements, Nazif said.

The results of these two new 10-year analyses, especially the P2S3i findings, “are reassuring—particularly with respect to our older patients, those over the age of 80, and those at intermediate risk—that TAVR really is a very reasonable, durable solution for them,” Nazif said.

“The vast majority of these patients will receive a single valve that will last their lifetime,” he added. “And so, I think that what this does is reassure us that our practice in that patient population is appropriate, while also further whetting the appetite for the longer-term results of the low-risk trials to inform lifetime management decisions in younger, lower-risk patient populations, who may require reintervention in their lifetime.”

Limitations of Long-term Follow-up

There are methodological limitations that need to be considered when interpreting the results of these 10-year studies, Nazif said, noting that many patients did not re-consent to additional follow-up beyond 5 years, including a slightly greater proportion in the SAVR versus TAVI group.

Although the vital status sweep provided more complete data for all-cause mortality, that didn’t address the missing data for other endpoints, like reinterventions, he said.

“This emphasizes,” Nazif said, “that 5 years will not be adequate to understand bioprosthetic valve durability moving forward and that trials in this space, especially in lower-risk populations, really need to be designed with planned 10-year follow-up, potentially augmented by long-term, real-world registry surveillance.”

An accompanying special communication by Sreekanth Vemulapalli, MD (Duke Clinical Research Institute, Durham, NC), and colleagues addresses the challenges associated with getting 10-year results from transcatheter valve trials. “With the shortage of research coordinators and the difficulties of following trial participants for 10 years, there is an urgent need to validate the use of reusable, efficient, scalable data sources for trial follow-up and adopt standardized approaches to statistical challenges associated with reconsent and nonrandom missing data,” they write.

For these specific analyses, Borger underscored the difficulties introduced by the older age of the populations, which comes with a very high competing risk of death. “More than 90% of patients were dead 10 years after entering the study, and therefore it’s difficult to make conclusions about survival,” he said.

A second major issue involves the methodology behind the vital status sweep, “which is probably the best that the investigators could do” since the studies were initially planned to have only 5 years of follow-up, Borger said. But, he added, “we don’t know if this methodology is valid. It would be more likely to detect somebody that is dead than somebody that is alive using that methodology, and we need to look at it in other patient groups and other studies to see if it accurately reflects the data that you would get from an ongoing randomized trial.”

Considering these limitations, “I don’t see a lot of changes in our practice patterns based on these two trials,” Borger said, noting that TAVI is already the default strategy for patients with an intermediate risk of surgery in the US and Europe.

Looking for Answers on TAVI Durability

Forthcoming 10-year results from the low-risk trials are really what’s needed to make determinations about the long-term durability of TAVI valves, Borger said.

“There is, of course, a worrisome signal from the self-expanding valves from the Evolut Low Risk Trial,” he said, alluding to the higher reintervention risk that emerged with TAVI versus SAVR at 6 to 7 years of follow-up. “PARTNER 3 does not have a worrisome signal at this point.”

There are several potential issues that could affect the durability of TAVI valves, Borger indicated.

“You have to recognize that there are inherent limitations to TAVI, such as not removing the calcium, which may have implications for blood stasis and thrombosis or healing with paravalvular leaks, or even just the amount of space that’s available for a future valve-in-valve procedure,” he said. “But regardless [of whether] it’s a balloon-expandable or a self-expanding [device], we also have to recognize that there’s definitely more manipulation of the valve leaflets with TAVR than there is with SAVR.”

And that can affect long-term durability, he added.

“The heart valve companies have told us as surgeons for decades now, ‘Don’t touch the valve leaflets,’ because that can impact long-term durability. They know this from their pulse duplicators and other preclinical information. And yet we are crimping these valves, we are using balloons to expand them, and sometimes we even use larger balloons to overexpand these valves,” Borger said. “This is not biomechanically surprising to see that this may have some impact on long-term durability.”

Within 3 years, he said, the field will see 10-year results from the low-risk trials that will help determine whether the number of TAVIs will continue to increase relative to SAVR or whether “we have a possible real problem that is going to lead to the pendulum swinging in the other direction back towards SAVR.”