BACKGROUND: Atrial fibrillation (AF) is a major cause of cardio-embolism in patients with stroke and transient ischemic attack (TIA). Insertable cardiac monitors (ICM) make long-term monitoring for AF possible, but limited health care resources make patient selection important. AF is associated with atherosclerosis and markers of this could potentially be used to guide AF monitoring.
METHODS AND RESULTS: One-hundred fourteen TIA-patients without AF were thoroughly monitored for AF with ECG, 72-hour Holter monitoring and ICM with a median monitoring time of 2.2 years. Patients with AF (n = 18) were significantly older than patients without AF (age 71.1 versus 64.4 years, P = .008) but were otherwise similar in regards to comorbidities. AF patients had significantly thicker carotid intima-media and also more often presence of carotid plaques than patients without AF, but no difference was found after adjusting for age and sex. No difference in noncontrast cardiac CT calculated coronary artery calcium score was found between the 2 groups. Serum biomarkers did not differ between groups, except for brain natriuretic peptide (BNP), where patients with BNP in the upper tertile were more likely to have AF than patients with BNP in the lowest tertile, odds ratio 5.96 (95% confidence interval 1.04-34.07, P = .045).
CONCLUSIONS: Carotid intima-media thickness and coronary artery calcium score were poor predictors of AF in patients with TIA. Apart from BNP, the examined biomarkers (hs-CRP, MR-proADM, c-TnI, copeptin) had no predictive value, but larger scale studies are needed to confirm these findings.
|Tidsskrift||Journal of Stroke & Cerebrovascular Diseases|
|Status||Udgivet - apr. 2020|