4.3 AF rotors and the fibrotic tissue properties
The development and progression of atrial fibrosis are considered as the substrate for AF perpetuation. LGE-MRI has been developed to visualize and quantify the extent of atrial fibrosis. The DECAAF multicenter prospective study demonstrated that atrial fibrosis estimated by LGE-MRI independently predicted AF recurrence after catheter ablation.14 Furthermore, computer simulations demonstrated AF rotors located in boundary zones between fibrotic and non-fibrotic tissue.4 To quantitatively characterize the fibrosis spatial patterns in each computer simulation model, the fibrosis density and fibrosis entropy were calculated. The local fibrosis density value was calculated as the proportion of fibrotic elements among all elements within the surrounding sub-volume. The local fibrosis entropy in each element was calculated as the level of disorganization within the surrounding sub-volume, quantified based on the modified Shannon entropy. However, this has not been verified in clinical use. In this study, the incidence of NPAs was higher in fibrotic tissue areas with an LGE-entropy of ≧5.7 (G1 and G2) as compared to that with an LGE-entropy of <5.7 (G3 and G4). No NPAs could be found in dense fibrotic tissue areas with an LGE-volume ratio of >50% regardless of the LGE-entropy. This indicated that the NPAs could be associated with tissue properties with fibrotic heterogeneity (entropy) rather than the fibrotic volume. The dense fibrotic tissue areas with an LGE-volume ratio of >50% appeared to be scar without electrical activity. This result was consistent with the computer simulation model. We found that the atrial myocardium consisted of different fibrotic tissue properties. Although it was not statistically proven, we speculated that the myocardial fibrosis gradually progressed and the fibrotic heterogeneity increased, which could have provided heterogenous fibrotic tissue where the AF rotors were frequently found. Those fibrosis patterns might correspond to the patchy LGE sites on the LGE-MRI. Further, the fibrosis progression decreased the fibrotic heterogeneity and increased the fibrosis volume, which could have provided the homogenous fibrosis tissue, namely dense scar. This tissue property might play an important role as an obstacle for macro-reentrant atrial tachycardia but not AF. Of importance, we might have to pay attention to the early stage fibrotic tissue properties (G1) that could progress to patchy fibrotic tissue associated with AF rotors in the future.
AF recurrence was observed in the patients with the slow highest %NP values where direct RF applications were performed or with the high highest %NP values where direct RF applications were not performed. The tissue properties of NPAs with a low highest %NP were characterized as having a low LGE entropy and LGE-volume ratio, which was categorized as G3. We speculated that those NPAs were an unmatured AF substrate and the effect of an NPA-targeted ablation would be less at the NPAs with G3 tissue properties as compared to that with G1 or G2 tissue properties. As for the catheter intervention for NPAs, direct RF applications were recommended in the previous study. That was because the real AF drivers were considered to be contained in the NPAs where rotational activations were frequently observed.10 When considering the reconduction of the ablation line, a direct RF application at the NPA would be necessary even though that area was locked in the PVI or Box lesion line.