Background:                      

Clinically, it can be difficult to separate atrial fibrillation (AF) from acute or chronic coronary syndrome in an emergency care setting, as these conditions can lead to similar symptoms, display ST depression in the electrocardiogram and show elevated cardiac biomarkers. 

Purpose:

We set out to investigate the predictive value of high sensitive cardiac troponin T (hs cTnT) for significant coronary artery disease (sCAD) in patients with newly diagnosed AF and unknown coronary artery status.

Methods:

We included consecutive patients who were admitted to our Chest-Pain-Unit between January 2013 and October 2019 with AF. To be eligible for inclusion, patients had to have newly diagnosed AF, to be ≥18 years old, and undergone coronary angiography during the current hospitalization. For identification of predictors of hs cTnT elevation and sCAD we performed a univariate logistic regression analysis. We conducted propensity score matching to adjust for previously identified predictors of hs cTnT elevation. By receiver operating characteristic analysis and calculation of Youden Indices optimal troponin T thresholds have been calculated. 

Results:

Out of 2535 AF patients we identified 144 patients with first diagnosis of AF and unknown coronary artery status who underwent coronary catheterization. At baseline 104 out of 144 patients (72%) had an elevated hs cTnT above the 99^thpercentile. By univariate logistic

regression analysis we identified “reduced left ventricular systolic function” as the only independent predictor of hs cTnT elevation (HR: 1.99; 95%-CI: 1.04-3,78, p=0.036).

77 out of the included 144 patients undergoing cardiac catheterization had diagnosis of sCAD (54%) requiring revascularization. Patients with sCAD were significantly older, had more frequent diagnosis of arterial hypertension and chronic kidney disease, a higher CHA²DS²-

VASc score and higher baseline hour-, 1 hour- and ∆ hs cTnT levels. Univariate logistic regression analysis suggested concomitant diagnosis of arterial hypertension (HR: 4.33; 95%-CI: 1.05-17.80, p=0.042),the CHA₂DS₂-VASc score (HR: 1.50; 95%-CI: 1.01-2.24,

p=0.046), baseline hour hs cTnT (HR: 1.01; 95%-CI: 1.00-1.03, p=0.045) and ∆ hs cTnT (HR: 1.20; 95%-CI: 1.09-1.33, p<0.001) to be associated with sCAD. After propensity score matching for possible confounders of hs cTnT elevation. only higher hs cTnT levels retained significant differences between the matched “sCAD” (53 patients) - and “no sCAD” (53 patients) – group. Receiver operating characteristic curve analysis of the matched groups showed an area under the curveof 0.65 (95%-CI: 0.54-0.75, p=0.01) for baseline hour hs cTnT and 0.77 (95%-CI: 0.67-0.87, p<0.001) for ∆ hs cTnT. The Youden Index for baseline hour hs cTnT was 0.25 with a corresponding cut-off of 20.5 ng/pl and 0.52 for ∆ hs cTnT with a corresponding cut-off of 4.5 ng/pl (sensitivity:0.84; specificity:0.68). The calculated negative predictive value Wert for ∆ hs cTnT <5 ng/pl was 0.84.

Conclusions:

Our study suggests that the “1-hour-delta” of hs cTnT (∆ hs cTnT) might be suitable to identify patients with sCAD in patients with newly diagnosed symptomatic AF in an emergency care setting. Multicentric studies including a large number of patients are necessary to verify our results, and to determine the ∆ hs cTnT threshold with the highest negative predictive value to avoid excessive cardiac catheterization in this population.