Atrial fibrillation, thromboembolism, and haemostasis: causal or casual associations?

Atrial fibrillation is the most common cardiac arrhythmia (1, 2) and is associated with an increased risk of death (3). In addition to the hemodynamic consequences such as loss of atrial contraction and loss of the chronotropic response to increased circulatory demands, atrial fibrillation causes important morbidity and mortality due to thromboembolic events, mainly stroke and myocardial infarction (4). Although atrial fibrillation can be terminated in the vast majority of patients (5), its recurrence can often not effectively be prevented (6-8). In these patients, heart rate control and anticoagulation are primary therapeutic goals. Oral anticoagulation with warfarin or phenprocoumon prevents approximately two thirds of thromboembolic events in patients with atrial fibrillation (9, 4, 10). Apart from typical risk factors for atherosclerosis (10), it is so far difficult to identify patients at high risk for thromboembolic events. In this issue, Vene et al. report that high baseline D-dimer levels are associated with future cardiovascular events in a series of 113 patients with chronic atrial fibrillation (11). Oral anticoagulation with warfarin reduced D-dimer levels, but D-dimer levels remained elevated in patients with cardiovascular events when compared to patients without events. The study is based on the previous concept of a “pro-coagulatory state” in patients with atrial fibrillation (12, 13) that may be initiated by reduced blood flow in the paralyzed, fibrillating atrium. Elevated D-dimer levels indicate increased fibrin formation and/or degradation and, hence, provide a biological marker for such a “pro-coagulatory state”. In addition, D-dimers may also be increased as part of an acute inflammatory response, and potentially also in extensive atherosclerotic disease. In view of this consideration, it is unfortunate that other markers of inflammation (CRP, white blood cell count) were not reported in the publication by Vene et al.(11). High D-dimer levels have been found in survivors of a stroke (14, 15), patients with atherosclerosis (16), and in patients with dementia (17). The present results confirm the suspicion that high fibrin turnover may be a contributing (risk) factor rather than a consequence of cardiovascular events in patients with atrial fibrillation. Whether this is due to the suspected pro-coagulatory effect of atrial fibrillation, due to accompanying inflammation, or due to extensive atherosclerosis, may be determined in future studies. The study by Vene et al. (11) also illustrates the advantages and disadvantages of a single centre study: Patient allocation and treatment was relatively uniform, but the small patient number required a combined end point without any effect on mortality reported. These limitations notwithstanding the data by Vene et al. (11) are a valid starting point for larger, prospective studies to determine the value of D-dimer levels for the prediction of cardiovascular events in patients with atrial fibrillation. Such studies will answer the open question whether D-dimer levels are a modifiable risk factor for thromboembolic events, as has been suggested by others (18), or an epiphenomenon of atherosclerosis and inflammation in patients with atrial fibrillation. In another article of the current issue, Poli and co-authors (19) published a study on the contribution of a specific genetic alteration to thromboembolic events in patients with atrial fibrillation. Due to the clinical observation that, despite a high risk for cerebral stroke, the majority of patients with atrial fibril© 2003 Schattauer GmbH, Stuttgart