Background: Although several classical studies seemed to provide clear ideas on the pathophysiology of atrial fibrillation, current concepts have to be modified on the basis of more recent findings.
Reentrant circuits: Based on the findings of Garrey and of Moe & Abildskov, atrial fibrillation has long been considered as the prototype of an arrhythmia being caused by multiple, random reentrant circuits, the number of which would determine the stability of the reentrant process. Local refractory and conduction properties would determine the size of individual circuits, a hypothesis quite convincing with respect to refractoriness, but so far hard to prove with respect to conduction. The finding that rapid atrial rates shorten atrial refractory periods and reverse rate adaptation (atrial remodeling) has coined the phrase "atrial fibrillation begets atrial fibrillation", indicating that any atrial tachyarrhythmia modifies the substrate in a way that favors reentry. With intracellular calcium overload being the initial trigger, down-regulation of genes encoding for calcium channels seems to primarily account for atrial remodeling. Primarily neglected concepts on the pathophysiology of atrial fibrillation suggesting single, meandering circuits or focal activity have regained attention. Atrial fibrillation as a random phenomenon is questioned not only by the dominant role of the left atrium for the maintenance of the arrhythmia, but also by most recent data demonstrating a spatio-temporal periodicity in activation patterns. Finally, ablation studies have provided convincing evidence that there is a subset of patients with focal or at least focally induced atrial fibrillation.