Current evidence primarily relies on observational studies without large randomized controlled trials (RCTs) directly comparing TAV-in-SAV and redo-SAVR. In a 2021 meta-analysis, Sá et al. [6] investigated observational studies published from 2015 to 2020 encompassing 16,207 patients with degenerated surgical aortic valves. The pooled analysis showed favorable 30-day all-cause mortality in patients treated with TAV-in-SAV compared to redo SAV (odds ratio [OR] and 95% confidence interval [CI]: 0.52 [0.39 to 0.68], p 0.001). However, the authors also reported higher odds of severe post-procedural patient-prosthesis mismatch (PPM) in the TAV-in-SAV group compared to the redo SAVR group (OR [95% CI]: 4.63 [3.05 to 7.03], p 0.001). However, the TAV-in-SAV group had lower odds of major bleeding (OR: 0.48 [0.28 to 0.80], p = 0.013) and shorter length of stay (absolute difference: –3.30 days [–4.52 to –2.08], p 0.001) compared to the redo SAVR group. No difference was observed in the rates of stroke, myocardial infarction, permanent pacemaker implantation, paravalvular leak (PVL), renal failure, and major vascular complications during a short-term follow-up.

Subsequently, Raschpichler et al. [7] expanded a meta-analysis to mid-term outcomes in patients undergoing TAV-in-SAV versus redo SAVR. The authors also demonstrated lower 30-day mortality in patients treated with TAV-in-SAV compared to redo SAVR (2.8% vs. 5.0%; risk ratio [RR] and [95% CI]: 0.55 [0.34 to 0.91], p = 0.02). However, no difference was found in two-year mortality rates (hazard ratio [HR] and [95% CI]: 1.27 [0.72 to 2.2], p = 0.37). They also showed a higher chance of prosthetic regurgitation (RR [95% CI]: 4.18 [1.88 to 9.3], p = 0.003) and severe PPM (3.12 [2.35 to 4.1]) in patients treated with TAV-in-SAV compared to redo SAVR. No significant difference was observed between groups for stroke, myocardial infarction, or permanent pacemaker implantation during a short-term follow-up.

In a 2023 meta-analysis, Formica et al. [8] confirmed a higher 30-day all-cause mortality in patients who underwent redo SAVR compared to TAV-in-SAV (HR [95% CI]: 2.12 [1.49 to 3.03], p 0.0001). The cumulative incidence of all-cause mortality at one year was 12.9% (95% CI: 11.3% to 14.6%) in the redo SAVR group compared to 9.9% (95% CI: 8.5% to 11.5%) in the TAV-in-SAV group. At five years, the cumulative incidence of all-cause mortality was 25.4% (95% CI: 22.6% to 28.6%) in the redo SAVR group and 27.6% (95% CI: 24.9% to 30.1%) in the TAV-in-SAV group. However, among patients who survived beyond the first year, a landmark analysis showed that redo SAVR was associated with a significantly lower risk of mortality compared to TAV-in-SAV (HR: 0.59; 95% CI: 0.44 to 0.79; p 0.001). These findings were consistent with the updated five-year meta-analysis by Sá et al. [9]. In their pooled analysis, the TAV-in-TAV group had lower 30-day mortality rates (OR [95% CI]: 0.52 [0.39 to 0.68], p 0.001), less major bleeding (OR [95% CI]: 0.48 [0.28 to 0.80], p = 0.013), but higher rates of severe PPM (OR [95% CI]: 4.63 [3.05 to 7.03], p 0.001). The authors also demonstrated that PPM became associated with all-cause mortality after eight months of follow-up. Their time-to-event analyses revealed that the relative risk between ViV TAVR and redo SAVR changes over time, indicating the presence of time-varying treatment effects.

These findings were recently reproduced by Awtry et al. [10], who investigated the five-year outcomes of Medicare beneficiaries who underwent either TAV-in-SAV or redo SAVR for a degenerated surgical aortic valve. In their propensity score–matched cohort analysis, patients who underwent redo SAVR had higher rates of 30-day mortality (9.6% vs. 5.0%, p 0.001), permanent pacemaker implantation (11.7% vs. 6.9%, p 0.001), and renal failure (12.9% vs. 3.6%, p 0.001). However, their landmark analysis demonstrated a survival benefit of redo SAVR over TAV-in-SAV for patients who survived the first 20 months and remained stable at five years (62.3% vs. 46.7%, p 0.001).

The Global Valve-in-Valve Registry represents the first large-scale assessment of the early experience with TAV-in-SAV in patients with degenerated surgical bioprostheses deemed unsuitable for redo surgery [11]. This registry demonstrated a marked improvement in functional status, with the proportion of patients in New York Heart Association (NYHA) class III or IV decreasing from 94% at baseline to 7.4% at 30 days post-procedure, which sustained through one-year follow-up. Similarly, other single-center studies demonstrated improved NYHA class following TAV-in-SAV [12, 13]. However, when compared directly, the magnitude of NYHA class improvement did not differ between patients undergoing TAV-in-SAV and those receiving redo SAVR.

The PARTNER 2 ViV registry data also showed improved NYHA functional class from baseline to 30 days and one year in high-risk patients who underwent a balloon-expandable TAV-in-TAV procedure (NYHA III/IV 90% at baseline, 45% at 30 days, and 43% at one year) [14]. Both the mean Kansas City Cardiomyopathy Questionnaire (KCCQ) score improved from 43 at baseline to 76 points at one year, and the mean 6-minute walk test from 164 m at baseline to 248 m at one year (both p 0.0001). Reported quality of life measured by the KCCQ score remained stable in survivors at five-year follow-up [15]. Recent registry data on the SAPIEN 3 valve indicate that TAV-in-SAV provides comparable improvements in functional status and quality of life from baseline to one year, similar to outcomes observed after TAVR for native aortic valve stenosis [16]. Similarly, the CoreValve U.S. Expanded Use Study demonstrated a reduction in NYHA class III/IV from 88% at baseline to 6% at one year, with a mean improvement of 29 points in KCCQ score over the same period in favor of self-expanding TAV-in-SAV (both p 0.0001). Notably, this functional improvement also remained at five years [17].

So far, published studies have consistently shown the safety and efficacy of TAV-in-SAV for failed surgical aortic bioprosthesis, with an early mortality benefit compared to redo SAVR in high-risk patients. The long-term durability and mortality differences between TAV-in-SAV and redo-SAVR remain in favor of redo SAVR if such a procedure is feasible. However, it is important to note that the observational design of published studies is subject to significant limitations, including selection bias, heterogeneity, and residual confounding, even with the use of propensity score adjustment.

Patients undergoing TAV-in-SAV often differed significantly from those receiving redo-SAVR in terms of higher surgical risk, older age, worse functional status, and more comorbidities. Published data were also collected over prolonged periods during which TAVR technology evolved substantially, limiting the validity of direct comparisons with redo SAVR. Some analyses also violate the assumption of proportional hazards, indicating the presence of time-varying treatment effects. Finally, most studies use data from high-volume centers or national registries, which may not reflect outcomes in community-based settings.

Another important consideration in the current literature is the broad application of the term “ViV TAVR”, which is often used to describe both transcatheter aortic valve-in-valve procedures performed within a TAV-in-TAV and those performed within a TAV-in-SAV. However, whether these represent distinct clinical and procedural entities remains an area of active investigation. In the Redo-TAVR registry, Landes et al. [18] evaluated outcomes in a propensity score–matched cohort undergoing TAV-in-TAV versus TAV-in-SAV. The authors reported a higher procedural success rate in TAV-in-TAV patients compared to TAV-in-SAV patients (73% vs. 62%, p = 0.045). They defined procedural success as a 30-day composite, including freedom from mortality, device-related intervention, major vascular or cardiac complications, a residual mean gradient of less than 20 mmHg, and less than moderate aortic regurgitation. Notably, the authors reported similar 30-day and 1-year mortality rates between TAV-in-TAV and TAV-in-SAV patients (30-day: 3% vs. 4%, p = 0.57, and 1-year: 12% vs. 10%, p = 0.63). In the TAV-in-TAV group, procedural complication rates were low, with 1.4% of patients having a stroke, 3.3% having valve malposition, 0.9% having coronary obstruction, and 9.6% needing a new permanent pacemaker. No difference was found in procedural safety, defined as a 30-day composite of all-cause mortality, stroke, major bleeding, major vascular complications, coronary obstruction, annular rupture, cardiac tamponade, acute kidney injury, at least moderate aortic regurgitation, need for permanent pacemaker, and device-related intervention (TAV-in-TAV: 70% vs. TAV-in-SAV: 72%, p = 0.71).

Similarly, the TRANSIT registry reported a VARC-2 device success rate in TAV-in-TAV patients of 79% with 33% of patients having stenotic degenerated TAV, 56% of them had regurgitant TAV, and 11% of them had a combined degeneration as the indication for ViV TAVR. The authors reported a 30-day mortality rate of 2.9% and a 10% mortality rate at one year. No cases of valve thrombosis or coronary obstruction were recorded [19]. Moreover, the TVT registry of self-expanding TAV-in-TAV procedures reported a VARC-2 device success rate of 95% [20]. Of those, 44% had a stenotic degenerated TAV and 62% had a regurgitant TAV as the reason for ViV TAVR. The 30-day rates of stroke were 3.1%, major vascular complications were 2.1%, major bleeding was 9.1%, and the need for PPM was 5.6%. There were no cases of coronary obstruction or myocardial infarction. The authors reported a 30-day mortality rate of 3.2% and a 17.7% mortality rate at one year. The most recent TVT registry analysis comparing balloon-expandable TAV-in-TAV with native TAVR found comparable outcomes [21]. Coronary obstruction occurred in 0.3%, device embolization in 0.1%, need for a pacemaker in 6.1%, and surgical conversion in 0.5%. No significant differences were observed between groups in 30-day (4.7% vs. 4.0%, p = 0.36) or 1-year mortality (17.5% vs. 19.0%, p = 0.57), stroke rates at 30 days (2.0% vs. 1.9%, p = 0.84) and 1 year (3.2% vs. 3.5%, p = 0.80), or device thrombosis (1-year: 0.6% vs. 0.1%, p = 0.09).

Although TAV-in-TAV is an attractive alternative to redo SAVR, not all patients can undergo TAV-in-TAV and may require TAVR explant with redo SAVR. In the EXPLANTORREDO-TAVR registry, Tang et al. [22] compared the outcomes of patients who underwent TAV-in-TAV with those who underwent TAVR explant with redo SAVR. Patients with TAVR-explant had more preexisting severe PPM (17.1% vs. 0.5%, p 0.001), whereas redo-TAVR patients had more structural valve deterioration as the indication for reintervention (63.7% vs. 51.9%, p = 0.023). Compared with TAV-in-TAV, TAVR-explant had higher cumulative mortality at 30 days (13.6% vs. 3.4%, p 0.001) and one-year follow-up (32.4% vs. 15.4%, p = 0.001). Using landmark analysis, authors reported that TAVR-explant patients who survived the first 30 days had similar mortality rates to TAV-in-TAV patients at one-year follow-up.

In the EXPLANT TAVR registry, Bapat et al. [23] reported similar mortality rates of 13.1% at 30 days and 28.5% at one year in TAVR explant patients. Indications for explanation included endocarditis (43.1%), structural valve deterioration (20.1%), PVL (18.2%), and PPM (10.8%). In the STS database, Fukuhara et al. [24] found a 30-day mortality rate of 18% among patients who underwent TAVR explant, with no difference between balloon-expandable and self-expandable devices. However, their findings were notable for higher rates of ascending aortic replacement with self-expandable valve explants compared to balloon-expandable valves (22% vs. 9%, p 0.001), but similar aortic root replacement (19% vs. 24%, p = 0.22). In a meta-analysis by Yokoyama et al. [25], patients who underwent TAVR explant had a 30-day mortality rate of 16.7% (95% CI: 12.2% to 21.2%), with endocarditis, structural valve deterioration, and PVL as the most common reasons for TAVR explantation.

The Redo-TAVR registry showed marked improvement in NYHA class III or IV from 80% at baseline to 10% at 30 days and 13% at 1 year following the TAV-in-TAV procedure [18]. The TRANSIT registry reported similar findings showing improvement in NYHA class III or IV from 74% at baseline to 7% at 30 days and 13% at one year [19]. These results were aligned with findings from the most recent TVT registry of balloon-expandable valves showing sustained improvement both in NYHA class III/IV and KCCQ score from baseline through the first year in patients treated with TAV-in-TAV (NYHA class III/IV: 82% at baseline vs. 14%, KCCQ: mean of 39 at baseline vs. 74 at one year) [21]. The TVT registry of self-expanding valves demonstrated similar improvement in patients’ functional status following TAV-in-TAV procedure (NYHA III/IV class: 62% at baseline vs. 27% at 30 days vs. 25% at one year; KCCQ: mean 38 at baseline vs. 71 at 30 days vs. 75 at one year, p 0.001).

It is important to interpret these findings cautiously, given the inherent limitations of the observational study designs that dominate the current TAV-in-TAV literature. These include small sample sizes, selection bias, patient heterogeneity, and residual confounding, even with the application of propensity score adjustments. It is also important to note that patients undergoing TAVR explant differ significantly from those receiving TAV-in-TAV. They often have prohibitive coronary anatomy, endocarditis, severe PPM, PVL, or require aortic or multivalvular surgery, all of which contribute to an increased risk of mortality.

Nonetheless, current observational data suggest that TAV-in-TAV offers similar short-term safety and mortality outcomes compared to TAV-in-SAV. However, TAV-in-TAV is also not suitable for all patients, and those requiring TAVR explantation with redo SAVR face substantially higher early mortality. Finally, long-term outcomes after TAV-in-TAV are missing and warrant further investigation.

When selecting between ViV TAVR and redo SAVR for patients with failed bioprosthetic valves, key determinants include patient age, surgical risk, and life expectancy. American and European guidelines emphasize the importance of comprehensive risk assessment using validated tools, such as the STS-PROM and EuroSCORE II, and evaluating frailty, comorbidities, and anatomical feasibility.

ViV TAVR is generally favored in older patients (typically $>$75–80 years) or those with elevated surgical risk or limited life expectancy, given its minimally invasive nature, reduced perioperative morbidity, and shorter recovery times. In contrast, redo SAVR is preferred in younger patients (typically 65–75 years) with life expectancy longer than one to two years and low surgical risk, due to the superior durability of surgical prostheses and more favorable long-term hemodynamics. For patients in the intermediate-risk category (ages 65–75), either approach may be appropriate, with the final decision informed by patient anatomy, valve characteristics, procedural feasibility, and patient preferences.

Despite growing observational evidence, randomized data directly comparing ViV TAVR and redo SAVR remain limited, underscoring the need for further trials to refine treatment algorithms. Nonetheless, shared decision-making, supported by a multidisciplinary Heart Team, will remain essential to balance short-term procedural safety with long-term valve durability and patient-centered outcomes.

The long-term management of bioprosthetic valves hinges on an understanding of the mode of their failure. Accurate classification informs procedural planning, guides valve selection, and helps to avoid suboptimal outcomes or unintended complications. The European Association of Percutaneous Cardiovascular Interventions (EAPCI) and, most recently, Valve Academic Research Consortium 3 (VARC-3) describe four potential mechanisms for failure of bioprosthetic aortic valves, including structural valve deterioration (SVD), non-structural valve dysfunction (NSVD), thrombosis, and endocarditis [26, 27]. Key mechanisms of SVD include leaflet disruption, wear and tear, leaflet fibrosis, and calcification. In contrast, NSVD mechanisms include PPM, pannus formation with leaflet entrapment, valve malposition, migration, late embolization, and PVL. Unlike SVD, which usually evolves gradually over time (except in cases of sudden leaflet tears), NSVD often appears at the time of the initial valve replacement and typically remains present during follow-up.

Multiple registries have identified severe PPM, particularly in aortic prostheses $\le$21 mm in diameter, as well as PVL, as strong predictors of poor prognosis [28, 29, 30]. Of note, PPM occurs more frequently with surgical than transcatheter aortic valve replacement, likely due to annular size constraints from residual calcification that limit the implantation of larger surgical valves. In contrast, transcatheter heart valves are designed to expand within the native annulus, often accommodating larger or more optimal sizing [29, 31]. Evidence also suggests that in ViV procedures, patients with pre-existing PPM or small surgical bioprostheses achieve lower postprocedural gradients when treated with supra-annular self-expanding valves compared to intra-annular balloon-expandable valves [28, 31, 32]. However, pre-existing severe PPM and PVL may not be adequately addressed with supra-annular ViV alone, particularly in cases where surgical aortic root enlargement or replacement would otherwise be required [22, 23]. In such scenarios, intentional valve fracture using high-pressure balloon inflation may be considered to improve hemodynamics. Importantly, the feasibility and safety of this approach depend on whether the degenerated valve is surgical or transcatheter. While balloon valve fracture is established in treating PPM in failed surgical bioprostheses, intentional overexpansion of transcatheter valves carries increased risks, including acute leaflet dysfunction, annular rupture, and may compromise valve durability [33].

Active infective endocarditis represents a fundamentally different clinical scenario and poses a major contraindication to ViV TAVR. Active endocarditis typically necessitates surgical valve explantation and debridement due to the presence of infection, tissue destruction, and abscess formation, all of which cannot be addressed by transcatheter means. As such, active endocarditis remains an absolute contraindication to ViV TAVR [4]. Nonetheless, in rare and highly selected cases, ViV TAVR has been attempted as a bailout strategy in patients presenting with acute decompensated heart failure due to bioprosthetic valve failure in the setting of treated endocarditis who were deemed prohibitive surgical candidates [34]. Such decisions must be made on a case-by-case basis, ideally through a multidisciplinary heart team discussion, and are not representative of standard practice.

Despite ViV TAVR being an attractive alternative to redo SAVR, certain anatomical and clinical factors often necessitate TAVR-explant instead of ViV. These challenges include low or obstructed coronary ostia, valve malposition or migration, small annulus, annular rupture, risk of mitral valve impingement, or inadequate vascular access, making transcatheter reintervention unfeasible [23]. In contrast, severe calcification of the ascending aorta (“porcelain” aorta), prior chest radiation, and prior cardiac surgery with at-risk coronary grafts would favor TAVR over redo SAVR [4]. The need for concomitant ascending aortic or multi-valvular surgery and coronary artery disease unsuitable for percutaneous treatment are also contraindications to ViV TAVR.