Simultaneous epicardial and endocardial activation sequence mapping in the isolated canine right atrium

Since the atria are thin-walled structures, most studies that have examined the spread of activation in the atria have assumed that they behave electrophysiologically as a two-dimensional surface. It was the objective of this study to determine whether or not this assumption is true by simultaneously mapping the epicardial and endocardial activation sequences in the right atrium.

Identical precisely superpositioned epicardial and endocardial electrode templates with 250 unipolar electrodes each were used to map the isolated canine right atrium (n = 8) during continuous perfusion and superfusion with Krebs-Henseleit buffer. Data were recorded during control conditions (normal sinus rhythm), continuous pacing (S1S1 = 300 msec), and premature stimulation (S1S2 = effective refractory period + 5 msec). Pacing was performed at two sites, one located on the inferior crista terminalis and one lateral to the crista terminalis on a pectinate muscle. Tachyarrhythmias were induced by a single extrastimulus during the continuous perfusion of acetylcholine (10(-3.5) mol/L). Individual electrode sites were correlated with the gross anatomy and histology. Activation time differences were calculated between each two corresponding epicardial and endocardial sites. There were differences in the activation times between the epicardium and endocardium during all experimental conditions. However, the average difference for each condition was < 1 msec, suggesting that overall activation did not spread faster on either the epicardium or the endocardium, even though in certain regions one surface could lead the other. The dispersion of time differences was smallest during normal sinus rhythm and continuous pacing (SD = 5.6-5.8 msec) and largest after premature stimulation (SD = 6.3 msec for crista pacing, p < 0.05; SD = 8.1 msec for pacing lateral to the crista, p < 0.001). Differences in the activation sequence correlated with the underlying anatomic architecture. The largest differences in activation times between the epicardium and endocardium were associated with those regions of the atrium where pectinate muscles ran below the epicardial surface. The pectinate muscles in those areas were often discontinuous with the epicardial surface and facilitated the discordant epicardial-endocardial activation. The discordant activation was also found in regions where the atrial wall thickness was < 0.5 mm and correlated with transmural differences in fiber orientation. A tachyarrhythmia induced in the presence of acetylcholine, which demonstrated a focal activation pattern, was shown to have a reentrant loop that used free-running muscle bundles connecting the epicardial and endocardial surfaces, resulting in a three-dimensional pathway.

The findings of this study demonstrate that epicardial and endocardial activation can be discordant in specific regions and that discordance increases with abnormal activation sequences. Many of the differences in the epicardial and endocardial activation can be correlated with the heterogeneity of the anatomic architecture of the right atrium. The study also demonstrates that reentry can occur in a three-dimensional plane using the epicardial and endocardial surfaces connected by transmural muscle fibers.