Spatial and temporal linking of epicardial activation directions during ventricular fibrillation in dogs. Evidence for underlying organization.

BACKGROUND: It remains controversial as to whether electrical activation during ventricular fibrillation (VF) is organized. To detect the presence of organization in VF, the direction of epicardial activation (EA) at multiple sites was examined by using vector mapping. If VF is not a random process, EA direction at a given site should be related to adjacent sites and prior beats. METHODS AND RESULTS: Thirteen dogs with healing myocardial infarction (MI) and four dogs without MI had VF induced by programmed stimulation. Using a plaque electrode array with a 2.5-mm interelectrode distance, 91 vector loops were created for each "beat" of VF. Direction of maximum EA was determined at each site for the first 10 consecutive beats of VF and for 10 consecutive beats recorded 5 seconds after VF was established. Spatial and temporal linking of EA directions was evaluated by the ability of activation direction at a given site to be predicted by activation directions at eight adjacent sites for the index beat and at eight adjacent sites and the site of interest for the preceding beat using stepwise linear regression. The strength of the model as reflected by the correlation coefficient (r) indicated the degree of linking. We determined 1) changes in the degree of linking over time during a given episode of VF (using a paired-difference t test), 2) differences in the degree of linking between the anterior and posterolateral walls in animals with (n = 4) and without (n = 4) MI (using two-way ANOVA), and 3) the effect of repeated inductions (n = 10) on the degree of linking (using one-way ANOVA with repeated measures). During 57 episodes of VF, r for each model ranged from 0.64 to 0.88 during the transition to VF to 0.39-0.78 during established VF (p < 0.0001 for the difference). The presence of MI, the site of recording, and repeated inductions did not affect the degree of linking. For each episode, spatial linking was more prominent than temporal linking. CONCLUSIONS: Electrical activation during VF is organized. The degree of linking of EA directions during VF is not affected by the presence of MI, the site of recording, or repeated inductions of VF. During the first 5 seconds of VF, the degree of linking decreases.