Silent brain infarcts after major cardiac interventions: a retrospective paired cohort study using high-resolution DW-MRI

BACKGROUND: Silent brain infarcts (SBI) after cardiac procedures are under-recognized. In this retrospective single-center cohort study, we used paired high-resolution diffusion-weighted MRI (HR-DWI) to quantify SBI incidence and lesion burden after transcatheter aortic valve replacement (TAVR) and atrial fibrillation (AF) ablation, and to explore patient-level and peri-procedural correlates of SBI. METHODS: This retrospective hospital-based cohort study was conducted was conducted at a cardiovascular referral center iin Kunming, China, between March 2024 and March 2025. We retrospectively included adults who underwent transcatheter aortic valve replacement (TAVR) or atrial fibrillation (AF) ablation and had paired 3-T high-resolution diffusion-weighted MRI (HR-DWI) within 3 days before and within 7 days after the procedure. SBI was defined as a new postoperative DWI lesion not present on baseline MRI and without clinical stroke. Logistic regression was used to assess SBI presence, and negative binomial regression was used for lesion count, while multivariable linear regression was used to evaluate mean lesion diameter and total lesion volume. RESULT: Among 128 patients (34 TAVR; 94 AF ablation; mean age 63 ± 9; 71% male), SBI occurred in 73 (57.0%), with 652 lesions identified. SBI incidence was higher after TAVR than after AF ablation (79.4% vs. 48.9%, p = 0.002). In multivariable logistic regression, older age (adjusted OR 1.087 per year, 95% CI 1.037-1.140), alcohol consumption (adjusted OR 3.343, 95% CI 1.036-10.787), NT-proBNP > 450 ng/L (adjusted OR 1.364, 95% CI 1.158-1.837), and larger left atrial diameter (adjusted OR 1.093 per mm, 95% CI 1.027-1.187) were independently associated with SBI occurrence. For lesion burden, longer procedure time was associated with higher lesion diameter (+ 0.027 mm/min, 95% CI 0.008-0.046) and total volume (+ 1.792 mm³/min, 95% CI 0.643-2.94) and with higher lesion count (adjusted IRR 1.009 per min, 95% CI 1.000-1.019). Intraoperative cardioversion increased mean diameter (+ 2.826 mm, 95% CI 1.567-4.085) and total volume (+ 193.978 mm³, 95% CI 117.444-270.513). Alcohol consumption was also associated with a higher lesion count (+ 12.05, 95% CI 1.29-22.81). CONCLUSIONS: Paired HR-DWI/MRI demonstrates a substantial burden of clinically silent cerebral injury after major cardiac interventions. Older age, alcohol consumption, elevated NT-proBNP, and larger left atrial diameter were associated with SBI occurrence, while longer operation time and intraprocedural electrical cardioversion were associated with greater lesion burden.