Differential Predictors of Early and Mid-to-Long Term Cerebrovascular Events Following Transcatheter Aortic Valve Implantation.

Background and Objectives: Cerebrovascular events remain among the most feared complications after transcatheter aortic valve implantation (TAVI). Whether early and post-discharge cerebrovascular events are associated with distinct clinical, echocardiographic, haemodynamic, and procedural predictors remains incompletely clarified. Materials and Methods: This retrospective, observational, two-center cohort study included 556 patients who underwent TAVI for severe symptomatic aortic stenosis between October 2011 and May 2023. The primary outcomes were in-hospital stroke and mid-to-long term stroke occurring after hospital discharge. Stroke diagnoses were established by neurological assessment and confirmed by cranial computed tomography and/or magnetic resonance imaging. Univariable comparisons and Firth penalized multivariable logistic regression analyses were performed to identify predictors of each outcome. Results: In-hospital stroke occurred in 14 patients (2.5%), whereas mid-to-long term stroke occurred in 16 patients (2.9%). In Firth penalized multivariable analysis, baseline mitral regurgitation grade (odds ratio [OR] 2.174, 95% confidence interval [CI] 1.299-3.638; p = 0.003), HbA1c (OR 1.644, 95% CI 1.111-2.433; p = 0.013), and prior cerebrovascular event (OR 4.269, 95% CI 1.150-15.846; p = 0.030) were independently associated with in-hospital stroke. For mid-to-long term stroke, post-dilatation (OR 9.862, 95% CI 2.555-38.067; p < 0.001) and very severe aortic stenosis phenotype (OR 3.649, 95% CI 1.343-9.916; p = 0.011) were independently associated with the outcome. Conclusions: Early and post-discharge cerebrovascular events after TAVI were infrequent but associated with distinct predictor profiles. In-hospital stroke was associated with baseline mitral regurgitation grade, HbA1c, and prior cerebrovascular event, whereas mid-to-long term stroke was associated with post-dilatation and very severe aortic stenosis phenotype. Given the limited number of events, these findings should be interpreted as exploratory and require validation in larger prospective cohorts.