Stroke after transcatheter aortic valve implantation: incidence, temporal trends and predictors.

BACKGROUND

Stroke remains one of the most serious complications of transcatheter aortic valve implantation (TAVI). This study evaluated temporal trends in 30-day stroke after TAVI, explored potential contributors to these trends and examined associations with mortality.

METHODS

All patients undergoing TAVI in Sweden between 2008 and 2023 were identified from the SWEDEHEART registry. The primary endpoint was 30-day ischaemic or haemorrhagic stroke. Temporal trends were analysed using Bayesian binomial regression. To investigate potential contributors to the observed trends, sequential models were fitted, adjusting for patient case-mix and centre-level procedural volume. Predictors of 30-day stroke were explored using a hierarchical Bayesian logistic regression with a centre-level random intercept. Mortality was assessed using Kaplan-Meier estimates and a multivariable Bayesian Cox model with a 30-day landmark.

RESULTS

Among 11 957 patients, 374 (3.1%) experienced a stroke within 30 days and 310 during index hospitalisation. The incidence of 30-day stroke declined from 5.3% in 2008 to 3.2% in 2023 (OR 0.97 per year; 95% credible interval 0.95-1.00; probability of direction 98.2%). The temporal decline was partly explained by changes in patient case-mix, particularly declining EuroSCORE (European System for Cardiac Operative Risk Evaluation) II. Centre procedural volume did not modify temporal trends as this trend was similar across low-volume, medium-volume and high-volume centres. In exploratory analyses, prior stroke, female sex, porcelain aorta and self-expanding valves were associated with a higher probability of 30-day stroke, whereas TAVI-in-SAVR (transcatheter aortic valve implantation in surgical aortic valve replacement) procedures were associated with a lower stroke risk. Stroke within 30 days was associated with markedly higher 30-day and 1-year mortality.

CONCLUSIONS

In this nationwide cohort, the incidence of early stroke after TAVI declined over time, partly reflecting changes in patient selection towards lower-risk profiles, with additional improvements in TAVI practice likely contributing. Despite this, early stroke remains strongly associated with excess mortality, underscoring the need for continued refinement of procedural strategies and targeted risk mitigation.