Biomarker, Imaging, and Clinical Factors Associated With Overt and Covert Stroke in Patients With Atrial Fibrillation
Gian Marco De Marchis,Philipp Krisai,Laura Werlen,Tim Sinnecker,Stefanie Aeschbacher,Tolga D Dittrich,Alexandros A Polymeris,Michael Coslovksy,Manuel R Blum,Nicolas Rodondi,Tobias Reichlin,Giorgio Moschovitis,Jens Wuerfel,Philippe A Lyrer,Urs Fischer,David Conen,Peter Kastner,André Ziegler,Stefan Osswald,Michael Kühne,Leo H Bonati,Swiss-AF Investigators
DOI: https://doi.org/10.1161/STROKEAHA.123.043302
IF: 10.17
Stroke
Abstract:Background: Atrial fibrillation is a major risk factor for stroke and silent brain infarcts. We studied whether a multimodal approach offers additional insights to the CHA2DS2-VASc score in predicting stroke or new brain infarcts on magnetic resonance imaging (MRI) over a 2-year follow-up. Methods: Swiss-AF is a prospective, multicenter cohort study of patients with known atrial fibrillation. We included patients with available brain MRI both at enrollment and 2 years later. The dates of the baseline and follow-up visits ranged from March 2014 to November 2020. The primary outcome was assessed 2 years after baseline and was defined as a composite of clinically identified stroke or any new brain infarct on the 2-year MRI. We compared a multivariable logistic regression model including prespecified clinical, biomarker, and baseline MRI variables to the CHA2DS2-VASc score. Results: We included 1232 patients, 89.8% of them taking oral anticoagulants. The primary outcome occurred in 78 patients (6.3%). The following baseline variables were included in the final multivariate model and were significantly associated with the primary outcome: white matter lesion volume in milliliters (adjusted odds ratio [aOR], 1.91 [95% CI, 1.45-2.56]), NT-proBNP (N-terminal pro-B-type natriuretic peptide; aOR, 1.75 [95% CI, 1.20-2.63]), GDF-15 (growth differentiation factor-15; aOR, 1.68 [95% CI, 1.11-2.53]), serum creatinine (aOR, 1.50 [95% CI, 1.02-2.22]), IL (interleukin)-6 (aOR, 1.37 [95% CI, 1.00-1.86]), and hFABP (heart-type fatty acid-binding protein; aOR, 0.48 [95% CI, 0.31-0.73]). Overall performance and discrimination of the new model was superior to that of the CHA2DS2-VASc score (C statistic, 0.82 [95% CI, 0.77-0.87] versus 0.64 [95% CI, 0.58-0.70]). Conclusions: In patients with atrial fibrillation, a model incorporating white matter lesion volume on baseline MRI and selected blood markers yielded new insights on residual stroke risk despite a high proportion of patients on oral anticoagulants. This may be relevant to develop further preventive measures.