Atrial fibrillation (AF) is often seen in heart failure or mitral valve disease with left atrial enlargement, and it is not uncommon that sinus rhythm is restored when, after surgical repair of the mitral valve, the size of the atria diminishes. This clinical observation suggests that mechanoelectric feedback may play an important role in the occurrence of atrial fibrillation. Theoretical formulations and high-density mapping have shown that the electrophysiological mechanism responsible for the maintenance of AF is the irregular propagation through the atria of several wavefronts. Inhomogeneity in electrophysiological properties, excitation wavelength (refractory period X conduction velocity) and tissue mass are determinant factors for the initiation and persistence of re-entrant arrhythmias. According to this, atrial stretch may contribute to atrial fibrillation development by increasing the atrial size, by decreasing the refractory period and/or the conduction velocity and by increasing their spatial dispersion. Modulation of the atrial electrophysiological parameters by stretch has been documented in both experimental studies and humans. In humans, we have shown that atrial stretch caused by ventricular contraction modulates the cycle length of atrial flutter by affecting the refractory period and the conduction properties of the underlying re-entrant circuit (Ravelli et al. 1994). Numerous studies have demonstrated stretch-induced changes in atrial refractoriness, although in the intact atrium conflicting results exist.

We have studied the effects of stretch on atrial refractoriness and vulnerability to atrial fibrillation in the isolated rabbit heart (Ravelli & Allessie, 1997). Atrial dilatation was obtained by raising the level of an outflow cannula in the pulmonary artery after occlusion of the caval and pulmonary veins. Increasing atrial pressure resulted in a progressive shortening of the atrial refractory period and monophasic action potential duration. All these changes were completely reversible after release of the atrial stretch. Dilatation of the atria was a major determinant for the vulnerability to atrial fibrillation. The inducibility of AF increased with progressively higher pressure levels and was strictly dependent on the refractory period. Atrial fibrillation was not induced at refractory periods > 70 ms. For refractory periods shorter than 50 ms, the inducibility of atrial fibrillation increased to 80 %. Similar results have been recently obtained in humans (Tse et al. 2001).

Whereas atrial stretch favoured the induction of AF by a premature beat, lowering of the atrial pressure invariably terminated AF. We have been investigating the mechanical modulation of atrial fibrillation by stretch release. Lowering of the atrial pressure causes a lengthening of AF cycle length and an increase in the degree of organization of atrial fibrillation, which leads to the arrhythmia termination. A gradual prolongation of the refractory period together with the diminishment of atrial size causes a progressive decrease in the average number of wavelets, which increases the chance of AF interruption.

These results provide evidence for mechanoelectrical feedback as a potential mechanism for atrial re-entrant arrhythmias by critical shortening of atrial refractoriness. An increase of spatial heterogeneity in refractoriness and/or conduction by stretch may also facilitate wavefront fragmentations by creating regions of functional block. In addition to changes in the substrate, stretch-induced atrial triggers may play a role in atrial fibrillation genesis. Recently ithas been proposed that clinical atrial fibrillation may start by stretch-activated atrial premature beats originating from pulmonary veins. The experimental observation of stretch-induced afterdepolarizations producing triggered atrial arrhythmias may support this hypothesis (Nazir & Lab, 1996). In the development of chronic atrial fibrillation, structural changes caused by chronic stretch should also be considered in addition to stretch-induced electrophysiological changes.