Computational modeling provides mechanistic insights into TAVI deterioration, but current durability predictions remain exploratory due to a lack of calibrated surrogate-to-outcome validation.
Computational modelling offers potential for understanding TAVI durability and guiding patient-specific planning, but requires rigorous validation before clinical application.
Abstract As transcatheter aortic valve implantation (TAVI) expands to younger populations, durability has become a concern, requiring a lifetime rather than a single-procedure perspective. While clinical trials suggest comparable mid-term performance to surgical bioprostheses, data beyond ten years remain limited, particularly for bicuspid valves, valve-in-valve procedures, and complex anatomies. Computational modelling combines patient anatomy and device design in computer-based simulations to study valve performance under physiological loading. Applied to TAVI, these models can reproduce implantation, evaluate mechanical stresses, and simulate blood flow, providing mechanistic insights into deterioration processes, including altered leaflet loading, stent deformation, and thrombosis-prone flow. Although these simulations do not directly assess durability, they use surrogate metrics linked with these mechanisms, helping identify factors that may influence longevity and guide design and procedural refinements. Clinically, modelling could support patient-specific planning and reintervention strategies, informing decisions across the valve-replacement pathway, an important consideration as younger patients are likely to undergo multiple lifetime procedures. Integrating these tools into pre-procedural planning may help anticipate challenges such as coronary access, annular geometry, and redo feasibility. However, current studies report elements of verification and field-level validation, but none complete a pre-specified, calibrated surrogate-to-outcome validation with uncertainty/sensitivity analysis; thus, durability predictions remain exploratory. Progress needs transparent verification, field checks versus bench or imaging, surrogate calibration to data, outcome testing in independent cohorts, and routine uncertainty/sensitivity reporting, with close clinician-engineer collaboration. This review underscores the need for a multidisciplinary approach and provides a critical analysis of the available tools and their potential to advance long-term outcomes.
Rauseo et al. (2026) conducted a review in TAVI durability. Computational modelling was evaluated. Computational modeling provides mechanistic insights into TAVI deterioration, but current durability predictions remain exploratory due to a lack of calibrated surrogate-to-outcome validation.