Background: While surgical aortic valve replacement (SAVR) is recommended for patients <70 years with severe aortic stenosis (AS), predictors of valve failure and the timing of subsequent valve-in-valve transcatheter aortic valve implantation (ViV-TAVI) remain incompletely understood. Although younger age at SAVR has traditionally been assumed to confer a higher risk of earlier reintervention, the impact of patient age and other clinical factors on the interval to ViV-TAVI is unclear. Better characterization of determinants of early versus late ViV-TAVI after SAVR is essential for optimizing lifetime management strategies in AS patients.
Objectives: To identify patient-level characteristics associated with early versus late ViV-TAVI after SAVR for severe AS.
Methods: We performed a retrospective analysis of 138 consecutive patients who underwent SAVR (1999–2024) followed by ViV-TAVI (Fig.1A). Patients were stratified into quartiles based on the interval between SAVR and ViV-TAVI: <6.6 years (Q1, n=35), 6.6–9.8 years (Q2, n=34), 9.8–13.1 years (Q3, n=35), and >13.1 years (Q4, n=34) (Fig.1B). Baseline demographics, procedural success of ViV-TAVI and 1-year clinical outcomes according to VARC-3 definitions were compared across quartiles.
Results: Patients in Q1–Q4 underwent SAVR at median ages of 72.8, 70.1, 70.7, and 66.7 years, respectively (p<0.001) (Fig.1C). Additionally, age was negatively correlated with the SAVR-to-ViV interval (Spearman r=-0,368; P<0.0001) (Fig. 1D). The proportion of female patients was 51.4% (Q1), 61.8% (Q2), 40.0% (Q3), and 58.8% (Q4) (p=0.27). The predominant valve failure mode before SAVR was combined aortic regurgitation/stenosis (51.4%, 36.4%, 50.0%, and 50.0%; p=0.668). Bovine stented valves were most frequently used across quartiles (67.6%, 71.0%, 79.4%, and 71.9%; p=0.62), and median SAVR valve size was comparable (23 mm [IQR 21–25] in all groups; p>0.05). Baseline comorbidities, EuroSCORE II (8.5–10.9; p=0.86), STS score (4.0–8.1; p=0.51), pre-ViV NYHA class, and LV function were similar across quartiles. Median age at ViV increased progressively (77.5–83.3 years; p=0.042). Device success improved with longer SAVR–ViV intervals (62.9%, 67.6%, 74.3%, and 91.2%; p=0.043). Self-expandable valve use predominated (85.7%, 79.4%, 85.7%, and 85.3%; p=0.66), with similar ViV valve sizes (23–26 mm; p=0.63). At 1 year, rehospitalization for heart failure occurred in 2.9%, 0%, 5.7%, and 0% of patients in Q1–Q4 (p=0.373). Stroke was observed in 0%, 7.4%, 3.3%, and 4% (p=0.505), and repeat intervention in 0%, 3.7%, 0%, and 0% (p=0.36). One-year mortality was 16.7% in Q1, 6.7% in Q2, and 0% in Q3–Q4 (p=0.197). No significant differences were observed in procedural complications or mortality between quartiles.
Conclusions: In patients undergoing ViV-TAVI after SAVR, older age at index SAVR was associated with shorter intervals to ViV, challenging the notion that younger patients are at highest risk for earlier reintervention. One-year complication rates were low and comparable across quartiles, supporting the safety and efficacy of ViV-TAVI across all SAVR–ViV intervals. Interpretation is limited by the retrospective design and potential selection and survivorship biases.
