Editorial CommentIn this issue of the Journal, Chang et al. 1 report observations on the size and shape of the left atrial (LA) appendage and the LA itself in patients with paroxysmal atrial fibrillation (AF) undergoing pulmonary vein (PV) isolation. The estimated LA volume, mathematically calculated from the anteroposterior, superoinferior, and laterolateral LA diameters, is larger in patients with AF than in controls without atrial arrhythmias. The LA appendage is also larger and its opening to the LA is less eccentric (more "rounded") in patients with AF. After a mean follow-up of 20 months, they observed that in patients without AF recurrence, appendage size decreased and the opening of the appendage to the LA became more eccentric, while estimated LA volume also decreased, but did not return to normal. The authors interpret the data as showing decreased LA and appendage stretch and recovery of muscle function. In contrast with this favorable evolution, in patients with AF recurrence, LA size and appendage size increased, while the opening of the appendage became less eccentric. This interpretation lacks the support of dynamic studies; however, it is consistent with previous echocardiographic observations demonstrating return of appendage systolic flow after cardioversion. 2 Perhaps in favor Figure 1. On the left, coronal (frontal plane) MRI cut of the LA showing maximum left atrial diameters. On the right, transverse (horizontal plane) MRI cut of the heart showing LA diameters as usually measured by echocardiography.of the role of muscle function in shaping the appendage opening is the observation by the same group that the PV become more eccentric as they become more dilated. 3 This could mean that PV muscle content is incapable of shaping the PV by itself, and the eccentricity of the enlarged vein could be the result of compression by the dilating LA 4 in the narrow space of the posterior mediastinum.The potential importance of anatomic and functional changes of the LA appendage with AF is obvious in relation with thromboembolic phenomena, but these observations on LA changes can also help us understand the pathogenesis of AF. Several years ago an MRI study by the Taipei VA Hospital group 5 compared LA and PV diameters in patients with persistent and paroxymal AF and subjects without AF; they found PV dilatation and increased transverse (right-left) LA diameter both in persistent and paroxysmal AF. This suggests that LA dilatation could be a part of the AF substrate even in those cases considered of focal origin, although LA dilatation could also be attributed to the AF episodes. The paper published in this issue of the Journal 1 clearly demonstrates a partially reversible anatomic remodeling effect produced by recurrent AF episodes, confirming previous observations after electrical cardioversion 6,7 and PV isolation. 8 The fact that dilatation may not be totally reversible raises the possibility of atrial dilatation underlying and possibly causing AF, in a vicious circle where atrial stretch would lead to...