Search

Background: Atrial fibrillation (AF) detection and treatment are key elements to reduce recurrence risk in cryptogenic stroke (CS) with underlying arrhythmia. The purpose of the present study was to assess the predictors of AF in CS and the utility of existing AF-predicting scores in The Nordic Atrial Fibrillation and Stroke (NOR-FIB) Study. Method: The NOR-FIB study was an international prospective observational multicenter study designed to detect and quantify AF in CS and cryptogenic transient ischaemic attack (TIA) patients monitored by the insertable cardiac monitor (ICM), and to identify AF-predicting biomarkers. The utility of the following AF-predicting scores was tested: AS5F, Brown ESUS-AF, CHA2DS2-VASc, CHASE-LESS, HATCH, HAVOC, STAF and SURF. Results: In univariate analyses increasing age, hypertension, left ventricle hypertrophy, dyslipidaemia, antiarrhythmic drugs usage, valvular heart disease, and neuroimaging findings of stroke due to intracranial vessel occlusions and previous ischemic lesions were associated with a higher likelihood of detected AF. In multivariate analysis, age was the only independent predictor of AF. All the AF-predicting scores showed significantly higher score levels for AF than non-AF patients. The STAF and the SURF scores provided the highest sensitivity and negative predictive values, while the AS5F and SURF reached an area under the receiver operating curve (AUC) > 0.7. Conclusion: Clinical risk scores may guide a personalized evaluation approach in CS patients. Increasing awareness of the usage of available AF-predicting scores may optimize the arrhythmia detection pathway in stroke units.

Introduction: Secondary stroke prevention depends on proper identification of the underlying etiology and initiation of optimal treatment after the index event. The aim of the NOR-FIB study was to detect and quantify underlying atrial fibrillation (AF) in patients with cryptogenic stroke (CS) or transient ischaemic attack (TIA) using insertable cardiac monitor (ICM), to optimise secondary prevention, and to test the feasibility of ICM usage for stroke physicians. Patients and methods: Prospective observational international multicenter real-life study of CS and TIA patients monitored for 12 months with ICM (Reveal LINQ) for AF detection. Results: ICM insertion was performed in 91.5% by stroke physicians, within median 9 days after index event. Paroxysmal AF was diagnosed in 74 out of 259 patients (28.6%), detected early after ICM insertion (mean 48 ± 52 days) in 86.5% of patients. AF patients were older (72.6 vs 62.2; p < 0.001), had higher pre-stroke CHA₂DS₂-VASc score (median 3 vs 2; p < 0.001) and admission NIHSS (median 2 vs 1; p = 0.001); and more often hypertension (p = 0.045) and dyslipidaemia (p = 0.005) than non-AF patients. The arrhythmia was recurrent in 91.9% and asymptomatic in 93.2%. At 12-month follow-up anticoagulants usage was 97.3%. Discussion and conclusions: ICM was an effective tool for diagnosing underlying AF, capturing AF in 29% of the CS and TIA patients. AF was asymptomatic in most cases and would mainly have gone undiagnosed without ICM. The insertion and use of ICM was feasible for stroke physicians in stroke units.