• atrial fibrillation
• platelet activation
• haemorheology
• thrombogenesis

• ADP, adenosine diphosphate
• AF, atrial fibrillation
• ANF, atrial natriuretic factor
• ANOVA, analysis of variance
• β-TG, plasma β thromboglobulin
• CTAD, citrate, theophylline, adenosine, and dipyridamole
• ELISA, enzyme linked immunosorbent assay

It has previously been believed that atrial thrombus, which is the source of thromboembolism in atrial fibrillation (AF), is formed after 48 hours of onset of AF, but there are limited clinical and pathophysiological data to confirm this convincingly.¹ Therefore, oral anticoagulation treatment is recommended before and after cardioversion of AF of more than two days’ duration, to minimise thromboembolic complications.² In contrast, restoration of sinus rhythm in patients with AF lasting less than two days has previously been regarded as “safe”, and prolonged anticoagulation treatment (except for heparin) was not thought to be mandatory.² Despite these guidelines, many instances of thromboembolism still occur in acute AF and a better understanding of the pathophysiological processes leading the thromboembolism in acute AF is needed. Indeed, conversion to sinus rhythm may increase the thrombogenic milieu of the left atrium.^1,3 Furthermore, evidence that AF confers a hypercoagulable state is growing.⁴ However, it is unclear exactly how long after the commencement of AF this increased risk of thrombosis becomes apparent and therefore clinically important. For example, does it take hours, days or weeks for the hypercoagulable state to develop?

We hypothesised that platelet activation, haemorheology, and thrombogenesis occur within 48 hours of onset of AF, as a pathophysiological mechanism contributing to increased thromboembolism in this condition. To test this hypothesis, we measured indices of platelet activation (ex vivo platelet aggregation, as well as plasma soluble P-selectin and plasma β thromboglobulin (β-TG)), haemorheology (haematocrit and plasma viscosity), and thrombogenesis (fibrin D-dimer) in 31 patients with acute onset AF, who were compared with 93 patients with permanent AF and 31 healthy controls in sinus rhythm. Furthermore, 15 patients with acute AF from this same cohort were reanalysed for the above parameters after (spontaneous) reversion to sinus rhythm at outpatient follow …