RESUMO
BACKGROUND: There is growing evidence that chemokines are potentially important mediators of the pathogenesis of atherosclerotic disease. Major atherothrombotic complications, such as stroke and myocardial infarction, are common among atrial fibrillation (AF) patients. This increase in risk of adverse events may be predicted by a score based on the presence of certain clinical features of chronic heart failure, hypertension, age 75 years or greater, diabetes and stroke (the CHADS2 score). Our objective was to assess the prognostic value of plasma chemokines CCL2, CXCL4 and CX3CL1, and their relationship with the CHADS2 score, in AF patients. METHODS: Plasma CCL2, CXCL4 and CX3CL1 were measured in 441 patients (59% male, mean age 75 years, 12% paroxysmal, 99% on warfarin) with AF. Baseline clinical and demographic factors were used to define each subject's CHADS2 score. Patients were followed up for a mean 2.1 years, and major adverse cardiovascular and cerebrovascular events (MACCE) were sought, being the combination of cardiovascular death, acute coronary events, stroke and systemic embolism. RESULTS: Fifty-five of the AF patients suffered a MACCE (6% per year). Those in the lowest CX3CL1 quartile (≤ 0.24 ng/ml) had fewest MACCE (p = 0.02). In the Cox regression analysis, CX3CL1 levels >0.24 ng/ml (Hazard ratio 2.8, 95% CI 1.02-8.2, p = 0.045) and age (p = 0.042) were independently linked with adverse outcomes. The CX3CL1 levels rose directly with the CHADS2 risk score (p = 0.009). The addition of CX3CL1 did not significantly increased the discriminatory ability of the CHADS2 clinical factor-based risk stratification (c-index 0.60 for CHADS2 alone versus 0.67 for CHADS2 plus CX3CL1 >0.24 ng/ml, p = 0.1). Aspirin use was associated with lower levels of CX3CL1 (p = 0.0002) and diabetes with higher levels (p = 0.031). There was no association between CXCL4 and CCL2 plasma levels and outcomes. CONCLUSION: There is an independent association between low plasma CX3CL1 levels and low risk of major cardiovascular events in AF patients, as well as a linear association between CX3CL1 plasma levels and CHADS2-defined cardiovascular risk. The potential for CX3CL1 in refining risk stratification in AF patients merits consideration.