Association of sST2 and hs-CRP levels with new-onset atrial fibrillation in coronary artery disease.

BACKGROUND

The data on biomarkers as predictors of atrial fibrillation (AF) in patients with coronary artery disease (CAD) are limited.

METHODS

A total of 1946 patients with CAD were recruited to the ARTEMIS study. At baseline, the study patients underwent clinical and echocardiographic examinations and had laboratory tests. The patients (n=1710) with the information about the occurrence of new-onset AF during the follow-up were included in the present analysis.

RESULTS

During 5.7±1.5years of follow-up, 143 (8.4%) patients developed a new-onset AF. Higher values of soluble ST2 (sST2) (20.2±10.8 vs. 17.5±7.2ng/mL, p=0.005), high-sensitivity troponin T (hs-TnT) (11.9±10.2 vs. 10.3±8.3ng/L, p=0.005), high-sensitivity C-reactive protein (hs-CRP) (3.3±5.9 vs. 2.0±4.4mg/L, p<0.001) and brain natriuretic peptide (BNP) (85.6±77.5 vs. 64.9±73.5ng/L, p<0.001) had significant associations with the occurrence of new-onset AF. In the Cox clinical hazards model, higher age (p=0.004), greater weight (p=0.045), larger left atrial diameter (p=0.001), use of asthma/chronic obstructive pulmonary disease medication (p=0.001) and lack of cholesterol lowering medication (p=0.008) had a significant association with the increased risk of AF. When the biomarkers were tested in the Cox clinical hazards model, sST2 (HR=1.025, 95% CI=1.007-1.043, p=0.006) and hs-CRP (HR=1.027, 95% CI=1.008-1.047, p=0.006) retained their significant power in predicting AF.

CONCLUSION

A biomarker of fibrosis, sST2, and a biomarker of inflammation, hs-CRP, predict the risk of occurrence of new-onset AF in patients with CAD. These biomarkers contributed to the discrimination of the AF risk model, but did not improve it markedly.