Aims: Elevated levels of high-sensitivity cardiac troponin T (hsTnT) are associated with adverse outcomes in numerous patient populations. Their value in prediction of stroke risk in patients with atrial fibrillation (AF) is in debate.
Methods: The study population included 2898 consecutive patients presenting with AF to the emergency department of the Department of Cardiology, Heidelberg University Hospital. Associations between hsTnT and stroke risk were assessed using multivariable Cox regression.
Results: Elevated hsTnT levels (>14 ng/L) were associated with increased risk of stroke. Even after adjustment for various risk factors, elevated hsTnT remained independently associated with stroke risk in patients with AF, adjusted hazard ratio 2.35 [95% confidence interval (CI): 1.26-4.36] (P = 0.007). These results were consistent across important subgroups (age, renal function, ejection fraction, CHA2DS2-VASc score and main admission diagnosis). For hsTnT, area under the receiver-operating-characteristic curve (AUC) was 0.659 [95% CI: 0.575-0.742], compared to 0.610 [95% CI: 0.526-0.694] for the CHA2DS2-VASc score. Inclusion of hsTnT in the multivariable model for stroke risk prediction consisting of all variables of the CHA2DS2-VASc score was associated with a significant improvement of its discriminatory power.
Conclusion: Elevated hsTnT levels are significantly associated with higher risk of stroke and provide prognostic information independent of CHA2DS2-VASc score variables. Measurement of hsTnT may improve prediction of stroke risk in patients presenting to an emergency department with AF as compared to risk stratification based only on clinical variables.