Highlights
- The 10-year PARTNER 2A randomized follow-up found higher all-cause mortality after balloon-expandable SAPIEN XT TAVR than after surgery in intermediate-risk severe aortic stenosis, with the excess risk concentrated in transthoracic access.
- Aortic valve reintervention was substantially more frequent after TAVR than after surgery at 10 years, despite similar mean transprosthetic gradients among the small subset with echocardiographic follow-up.
- Contemporary 10-year data from the newer-generation SAPIEN 3 platform suggest that improved device design and more routine transfemoral implantation can narrow or eliminate the long-term gap versus surgery.
- Across the broader evidence base, long-term outcomes increasingly reflect access route, device generation, patient selection, and competing mortality rather than a simple TAVR-versus-SAVR dichotomy.
Background
Severe symptomatic aortic stenosis is a prototypical structural heart disease in which untreated obstruction carries high rates of heart failure, hospitalization, and death. Over the past decade, transcatheter aortic valve replacement (TAVR) has moved from a therapy reserved for prohibitive surgical risk to an established alternative to surgical aortic valve replacement (SAVR) across multiple risk strata. The clinical success of TAVR, however, has always depended on more than short-term procedural safety. For an elderly but still relatively active intermediate-risk population, the key questions are durability, late survival, valve reintervention, and the consequences of access route and prosthesis design over time.
The PARTNER 2A trial was one of the pivotal randomized studies that established TAVR as a viable treatment option for intermediate-risk patients. Yet the original follow-up horizon was not long enough to answer the durability and late comparative-effectiveness questions that now matter most to clinicians and health systems. The newly reported 10-year randomized outcomes of balloon-expandable SAPIEN XT TAVR versus SAVR, published in the Journal of the American College of Cardiology and indexed in PubMed (PMID: 42300821), therefore represent an important milestone in the long-term evidence base. They do not simply extend earlier data; they sharpen the field’s understanding of how device generation, access strategy, and patient trajectory influence outcomes over a decade.
Key Content
1. The PARTNER 2A 10-year randomized dataset: design and interpretive strengths
The PARTNER 2A long-term follow-up evaluated patients randomized between 2011 and 2013 at 57 centers to either balloon-expandable TAVR with the SAPIEN XT system or SAVR. Randomization was stratified by anatomical suitability for transfemoral (TF) versus transthoracic access, the latter combining transapical and transaortic (TA/TAo) approaches. This stratification is crucial: by 2011–2013, transfemoral access was increasingly preferred, but a substantial minority of patients still required transthoracic delivery, which is now rarely used in contemporary practice.
The analysis included 1,910 randomized patients who received a valve: 974 in the TAVR arm and 936 in the surgical arm. The cohort was elderly, with a mean age of 81.6 years, 45.4% women, and a mean Society of Thoracic Surgeons score of 5.8%, reflecting intermediate operative risk. Ten-year follow-up completeness was relatively strong for such a prolonged study: vital status was available for 90.5% of TAVR and 89.5% of surgery patients. Investigators also used vital status sweeps and required reconsent at 5 years to support long-term ascertainment.
These methodological details matter because long-horizon structural heart trials face predictable threats to validity: differential loss to follow-up, informative censoring, evolving practice patterns, and competing non-cardiovascular mortality in older adults. The PARTNER 2A report deserves credit for acknowledging these limitations and implementing strategies to preserve data completeness. Even so, as in most long-term valve trials, the certainty of echocardiographic comparisons is constrained by the small number of survivors available for imaging at 10 years.
2. Survival at 10 years: a signal favoring surgery, but mainly in transthoracic TAVR
The principal clinical finding is straightforward: all-cause 10-year mortality was 86.1% after TAVR versus 82.8% after SAVR, corresponding to a hazard ratio of 1.13 (95% CI, 1.02–1.25; P = 0.02). In absolute terms, both event rates are high, underscoring the advanced age and comorbidity burden of the study population; nonetheless, the excess mortality with TAVR was statistically significant.
The more nuanced and clinically important observation is the effect modification by access route. In the transfemoral subgroup, mortality was similar between TAVR and surgery (83.9% vs 82.1%; P = 0.27). In contrast, mortality was higher after TA/TAo TAVR than after surgery (93.2% vs 85.1%; P < 0.01), with a significant interaction by access route (P for interaction = 0.03). This pattern strongly suggests that the long-term disadvantage observed in the overall TAVR cohort was not an inherent property of transcatheter valve therapy per se, but rather a reflection of the then-common need for transthoracic access in anatomically challenging patients.
From a clinical standpoint, this distinction is critical. Modern TAVR practice has become overwhelmingly transfemoral, enabled by smaller delivery systems, refined imaging, and broader anatomic eligibility. Therefore, the trial’s overall mortality signal should not be generalized indiscriminately to contemporary transfemoral TAVR. Instead, the data reinforce a durable principle: access route remains a major determinant of outcome, and nontransfemoral TAVR should be considered a distinct procedural phenotype with different risk expectations.
3. Reintervention: a durability signal that remains relevant
At 10 years, cumulative aortic valve reintervention was 6.3% after TAVR versus 1.6% after surgery (P < 0.001). This difference is clinically meaningful, even though the absolute numbers remain modest. In an elderly cohort, late reintervention is uncommon because competing mortality is high; thus, a seemingly small absolute difference can still represent a relevant durability signal.
Several interpretations are possible. First, early-generation TAVR valves, especially in the SAPIEN XT era, were associated with higher rates of paravalvular regurgitation, annular sizing constraints, and structural valve deterioration compared with contemporary devices. Second, reintervention after TAVR may reflect both true bioprosthetic degeneration and residual hemodynamic or anatomical limitations that become clinically apparent over time. Third, the threshold for reintervention may differ between catheter-based and surgical patients because explanting a transcatheter valve can be technically complex and may occur later in the disease course.
The practical message for clinicians is not that all TAVR valves fail early, but that device generation matters. A decade ago, early balloon-expandable valves carried uncertainties that are progressively less applicable to today’s platforms. Still, durability remains an essential part of shared decision-making in intermediate-risk patients who may survive long enough to experience late valve failure.
4. Echocardiography at 10 years: similar gradients, but very limited ascertainment
Among the small number of patients with available 10-year echocardiographic data (24 TAVR and 35 surgical patients), mean gradients were nearly identical: 12.6 mm Hg after TAVR and 12.7 mm Hg after SAVR. This finding is reassuring at face value, suggesting preserved hemodynamics in surviving valves. However, the sample size is extremely small and almost certainly subject to survivor bias.
In other words, the echocardiographic subset represents a highly selected group of patients who survived long enough and remained available for imaging. These data can support feasibility and signal durability, but not definitive comparative claims about hemodynamic superiority at 10 years. This is an important limitation for readers to appreciate: in long-term valve studies, survival itself often determines who is still available for valve assessment, and this can obscure late structural differences.
5. How PARTNER 2A fits with the broader 2026 evidence landscape
The strongest way to interpret the new PARTNER 2A report is alongside contemporary long-term trials and synthesis studies, not in isolation.
The most relevant comparator is the 10-year PARTNER 2 SAPIEN 3 analysis (PMID: 42300820), which compared patients treated with the newer-generation SAPIEN 3 valve in the PARTNER 2 SAPIEN 3 Intermediate-risk Registry with surgery from PARTNER 2A using propensity score matching. In that study, 10-year mortality was essentially identical between TAVR and SAVR (83.4% vs 82.3%; HR 1.01; 95% CI, 0.91–1.13; P = 0.82), and reintervention rates were also similar (2.0% vs 1.9%). Mean gradients at 10 years were comparably low. Although not randomized head-to-head, this analysis suggests that the unfavorable long-term signal seen with SAPIEN XT may not extend to newer-generation balloon-expandable valves.
That distinction is highly clinically relevant. SAPIEN XT and SAPIEN 3 are not interchangeable technologies. The latter introduced enhanced sealing and design improvements that likely reduced paravalvular leak and improved procedural reproducibility, especially via transfemoral access. Thus, the PARTNER 2A 10-year results should be viewed as an important historical benchmark rather than the final word on modern TAVR durability.
The broader randomized evidence base also points in a similar direction. At 5 years, a 2026 Bayesian meta-analysis of six randomized trials involving 7,249 low- to intermediate-risk patients found higher all-cause mortality with TAVI than with SAVR (RR 1.12; 95% CrI 1.02–1.22) and a probable increase in stroke risk, with no evidence that surgical risk category modified the effect. A separate analysis of routine-practice versus trial SAVR cohorts raised external-validity concerns, showing that trial surgical controls may be more selected and have worse long-term survival than routine SAVR patients. These findings complicate interpretation: if RCT surgical arms are not fully representative of everyday surgery, then both the absolute and relative differences seen in trials may not perfectly generalize to broader clinical populations.
At the same time, the 7-year PARTNER 3 results in low-risk patients showed no significant difference in major composite outcomes between transfemoral TAVR and surgery, and the 5-year Evolut Low Risk data also showed similar death/stroke outcomes, albeit with higher pacemaker rates and some late reintervention signal in self-expanding systems. Taken together, the evidence indicates that long-term outcomes are highly device- and population-dependent. The field has moved from asking whether TAVR can work to asking which valve, in which anatomy, by which access route, for which patient, and over what expected lifespan.
6. Mechanistic and translational interpretation: why access route and device era matter
The PARTNER 2A findings are biologically plausible. Transthoracic TAVR, especially transapical implantation, was associated with more invasive access, myocardial injury risk, bleeding, and respiratory complications than transfemoral TAVR. In elderly intermediate-risk patients, these peri-procedural and subacute insults may compound frailty trajectories and contribute to late mortality. If a patient’s anatomy necessitates a transthoracic route, the “minimally invasive” benefit of TAVR diminishes substantially.
Device generation is also mechanistically important. Early balloon-expandable valves had less refined annular sealing and deployment control than current devices, increasing the likelihood of paravalvular leak and suboptimal expansion. Even if mean gradients remain acceptable among survivors, early prosthesis-patient mismatch or residual regurgitation can influence long-term ventricular remodeling and clinical outcomes. In contrast, modern transfemoral TAVR with improved devices may better preserve effective orifice area and reduce late valve-related events.
This is where echocardiography remains translationally valuable. Hemodynamic metrics such as mean gradient, effective orifice area, and paravalvular leak are not merely descriptive—they are the bridge between procedure and downstream biology. However, the long-term imaging evidence is now sparse because so few trial participants survive to be imaged a decade later. Future studies may need linked imaging registries and serial health-status phenotyping rather than relying solely on traditional trial follow-up.
7. Sex, anatomy, and subgroup considerations
The PARTNER 2A cohort was nearly half women, but the published 10-year summary does not indicate a major sex-specific differential effect. That said, sex-specific anatomy influences access feasibility, annular dimensions, and vascular complications, which may partially shape procedural selection and long-term outcomes. Other recent randomized work, including the DEDICATE-DZHK6 trial subgroup analysis in women and men, suggests that TAVI can achieve favorable early outcomes across sexes, though stroke and complication profiles differ by treatment and patient characteristics.
Similarly, anatomy remains decisive in bicuspid aortic stenosis and in patients with concomitant valvular disease. Recent observational data and meta-analyses suggest that TAVR outcomes in bicuspid valves may be less favorable in midterm follow-up than SAVR, likely because of asymmetric calcification, elliptical annuli, and device malexpansion. These anatomic complexities are conceptually related to the PARTNER 2A transthoracic subgroup: when anatomy is difficult, both procedural selection and long-term outcome can shift unfavorably.
8. Guideline and bedside implications
Current practice guidelines have increasingly favored TAVR in older patients with suitable anatomy, especially when transfemoral access is feasible. The PARTNER 2A 10-year data do not overturn that paradigm, but they do refine it.
For intermediate-risk symptomatic severe aortic stenosis, the evidence now supports a differentiated approach:
- Transfemoral TAVR remains highly attractive in appropriately selected older patients, especially with contemporary valves.
- Transthoracic TAVR should be reserved for carefully selected cases and recognized as a different risk category with less favorable long-term signals.
- SAVR remains a robust option for patients in whom durability, concomitant surgical needs, or anatomy favor surgery, particularly when life expectancy extends beyond the horizon of the current valve’s expected longevity.
In heart-team discussions, these data can be used to frame expectations realistically: not all TAVR is equivalent, and the expected lifetime of the patient should be matched to the expected durability of the prosthesis and access strategy.
Expert Commentary
The PARTNER 2A 10-year report is best understood as both a landmark and a cautionary tale. It is a landmark because few randomized valve trials achieve a decade of follow-up with vital-status completeness above 89%. It is a cautionary tale because the overall inferiority signal for SAPIEN XT TAVR should not be generalized to contemporary transfemoral TAVR with newer-generation devices.
Three major controversies emerge.
First, is surgery truly superior long-term in intermediate-risk patients? The answer is “not uniformly.” The disadvantage in PARTNER 2A was concentrated in transthoracic access, not transfemoral TAVR. Contemporary transfemoral TAVR with improved valves appears to close much of the gap.
Second, how much weight should be given to reintervention differences? In elderly patients, competing mortality reduces the absolute importance of valve reintervention, but durability still matters for quality of life, health-system utilization, and the subset of patients with longer survival expectancy. A 6.3% versus 1.6% reintervention difference over 10 years is not trivial, especially if future candidates are younger and more likely to outlive the first valve.
Third, how should long-term trial data influence guideline language? Guidelines increasingly emphasize patient-centered selection, anatomy, age, life expectancy, and local expertise. The new evidence supports that direction rather than a simplistic hierarchy. The durability and survival profile of a valve cannot be separated from access route, vascular anatomy, calcification burden, and prosthesis generation.
From a translational perspective, the field is entering a phase in which long-term success will depend on reducing paravalvular regurgitation, minimizing pacemaker need, improving leaflet durability, and expanding access to concomitant imaging and registry-based surveillance. The next generation of comparative studies should be designed to capture not just mortality, but functional status, frailty progression, structural valve deterioration, and the outcomes of valve explant or redo intervention.
Conclusion
The 10-year randomized PARTNER 2A results provide an important long-term benchmark for intermediate-risk severe aortic stenosis treated with balloon-expandable SAPIEN XT TAVR versus surgery. Overall mortality was higher with TAVR, and valve reintervention was significantly more frequent, but the excess mortality was driven mainly by transthoracic access rather than transfemoral implantation. Echocardiographic gradients among the small surviving subset were similar at 10 years, though imaging ascertainment was limited.
When interpreted alongside newer-generation SAPIEN 3 data and broader randomized meta-analyses, the message is nuanced: long-term TAVR outcomes are highly dependent on access route, device era, and patient anatomy. For contemporary practice, transfemoral TAVR in carefully selected intermediate-risk patients remains a compelling strategy, whereas older transthoracic TAVR experience should not be extrapolated to today’s procedures. The evolving evidence base supports a heart-team approach grounded in anatomy, life expectancy, and prosthesis durability rather than one-size-fits-all procedural preferences.
References
- Thourani VH, von Stein P, Mack MJ, et al; PARTNER 2 Investigators. 10-Year Randomized Outcomes of Transcatheter or Surgical Aortic Valve Replacement in Intermediate-Risk Aortic Stenosis. J Am Coll Cardiol. 2026;87(23):3309-3321. PMID: 42300821.
- 10-Year Outcomes of SAPIEN 3 Transcatheter Aortic Valve Replacement or Surgery in Intermediate-Risk Patients. J Am Coll Cardiol. 2026;87(23):3296-3308. PMID: 42300820.
- Updated 5-year outcomes of transcatheter versus surgical aortic valve replacement in patients with severe aortic stenosis at low- to intermediate-surgical risk. Heart. 2026. PMID: 41672766.
- Representativeness of surgical controls in aortic valve replacement trials: comparison with routine surgical cohorts. Heart. 2026;112(5):239-245. PMID: 40633940.
- Transcatheter or Surgical Aortic-Valve Replacement in Low-Risk Patients at 7 Years. N Engl J Med. 2026;394(8):773-783. PMID: 41144631.
- Six-Year Outcomes After Transcatheter vs Surgical Aortic Valve Replacement in Low-Risk Patients With Aortic Stenosis. J Am Coll Cardiol. 2026;87(23):3210-3221. PMID: 41697183.
- Sex-specific outcomes after transcatheter or surgical treatment of aortic valve stenosis: the DEDICATE-DZHK6 trial. Eur Heart J. 2026;47(11):1339-1353. PMID: 40900118.
- Echocardiographic Results of Transcatheter Versus Surgical Aortic Valve Replacement in Women With Severe Aortic Stenosis: The RHEIA Trial. J Am Heart Assoc. 2026;15(1):e047196. PMID: 41467363.
