Atrial fibrosis in embolic stroke of undetermined source: A multicenter study

BACKGROUND AND PURPOSE: Left atrial (LA) cardiac disease is a suspected cause of embolic stroke of undetermined source (ESUS). We tested the hypothesis that LA fibrosis, quantified using late-gadolinium-enhancement magnetic resonance imaging (LGE-MRI), predicts recurrent stroke or atrial fibrillation (AF) in patients with ESUS. METHODS: We compared atrial fibrosis in healthy controls and patients with lacunar stroke, ESUS, and known AF with or without prior stroke. We followed patients with ESUS prospectively for the primary outcome of recurrent ischemic stroke, incident AF, or both. RESULTS: We enrolled 203 patients from three centers: 103 patients without AF (35 healthy controls, 15 with lacunar strokes, 53 with ESUS) and 100 patients with AF (50 with and 50 without prior stroke). Patients with ESUS had significantly higher atrial fibrosis (15.0 ± 6.2%) compared to healthy controls (8.1 ± 7.9%; <0.0001) and compared to lacunar stroke patients (10.8 ± 8.4; p = 0.02), but had comparable fibrosis to patients with AF with (17.9 ± 11.4%) or without prior stroke (16.6 ± 9.2%; p = NS for both). Over a mean follow-up of 19 months, nine of 53 patients (16.9%) with ESUS experienced the combined primary outcome, which included six patients (11.3%) with recurrent ischemic stroke and five patients with incident AF (9.4%). Patients with ESUS with fibrosis ≥12% had a higher proportion of the combined outcome: 25.0% vs. 4.8%; p = 0.039. CONCLUSIONS: Patients with ESUS demonstrate atrial fibrosis comparable to that seen in AF. Atrial fibrosis ≥12% was associated with recurrent stroke, incident AF or both. This subgroup of ESUS patients may benefit from anticoagulation for secondary prevention of ischemic stroke.