Background and purpose: There are currently no biomarkers to select cryptogenic stroke (CS) patients for monitoring with insertable cardiac monitors (ICMs), the most effective tool for diagnosing atrial fibrillation (AF) in CS. The purpose of this study was to assess clinically available biomarkers as predictors of AF.
Methods: Eligible CS and cryptogenic transient ischaemic attack patients underwent 12-month monitoring with ICMs, clinical follow-up and biomarker sampling. Levels of cardiac and thromboembolic biomarkers, taken within 14 days from symptom onset, were compared between patients diagnosed with AF (n = 74) during monitoring and those without AF (n = 185). Receiver operating characteristic curves were created. Biomarkers reaching area under the receiver operating characteristic curve ≥ 0.7 were dichotomized by finding optimal cut-off values and were used in logistic regression establishing their predictive value for increased risk of AF in unadjusted and adjusted models.
Results: B-type natriuretic peptide (BNP), N-terminal pro-brain natriuretic peptide (NT-proBNP), creatine kinase, D-dimer and high-sensitivity cardiac troponin I and T were significantly higher in the AF than non-AF group. BNP and NT-proBNP reached the predefined area under the curve level, 0.755 and 0.725 respectively. Optimal cut-off values were 33.5 ng/l for BNP and 87 ng/l for NT-proBNP. Regression analysis showed that NT-proBNP was a predictor of AF in both unadjusted (odds ratio 7.72, 95% confidence interval 3.16-18.87) and age- and sex-adjusted models (odds ratio 4.82, 95% confidence interval 1.79-12.96).
Conclusion: Several clinically established biomarkers were associated with AF. NT-proBNP performed best as AF predictor and could be used for selecting patients for long-term monitoring with ICMs.
Keywords: atrial fibrillation; biomarkers; cardiac rhythm monitoring; cryptogenic stroke; insertable cardiac monitors.
© 2023 European Academy of Neurology.