The results described thus far suggest that PtAF induced
The results described thus far suggest that PtAF-induced atrial remodeling combined with acute atrial stretch facilitates anchoring of 3-D scroll waves to the border of atrial pectinate muscle and prolongs their life span in favor of AF maintenance. However, why is it that scroll waves anchor to the edges of thick muscle bundles such as pectinate muscles? To answer this question, we implemented a simplified 3-D numerical simulation of the human PtAF model using the equations of Courtemanche–Ramirez–Nattel et al. [60,61]. Two levels (high, 4μm; low, 2μm) of stretch-activated eletriptan were computed to mimic recordings documented by Kamkin et al. . In this simulation, the effects of stretch-activated channels, such as depolarization of the resting potential, were greater in the thin segments. Fig. 7A presents a diagram showing the spatial dimensions of the atrial model. Fig. 7B shows an example of an I-shaped filament (indicated by a red dotted line) scroll wave. I-shaped filament scroll waves induced by S1–S2 stimulation were mainly located at the junction between thin and thick myocardial segments. One theoretical explanation could be the “minimal length principle” described by Wellner et al.  which states that the filament of a scroll wave tends to drift to areas with conditions contributing to minimize its length. For instance, in thick myocardium, the conduction velocity is normal and scroll wave filaments are subject to positive tension, resulting in a strong tendency to drift toward the thinnest myocardial segments. Conversely, in thin myocardium, increased wall tension leads to a preferential activation of stretch-activated channels, leading to depressed conditions of impulse propagation. Consequently, scroll wave filaments in thin regions are exposed to negative tension and tend to drift toward the thickest myocardial segments. Such results suggest that scroll wave filaments are submitted to opposite drifting forces resulting from the heterogeneity of impulse velocities in thin and thick myocardium. Thus, as depicted in Fig. 7C, the balance between strong positive tension in thick regions and negative tension in thin regions could explain why I-filament scroll waves are preferentially located at the junction between thin and thick myocardium.
Fibrosis governs AF dynamics in failing hearts It is well known that heart failure increases the incidence of atrial tachyarrhythmias, resulting in a significantly higher rate of morbidity and mortality . Heart failure induced by rapid pacing leads to electrical as well as structural remodeling of atrial muscle, and the latter is characterized by extensive atrial interstitial fibrosis . Atrial interstitial fibrosis associated with heart failure may promote AF maintenance even in the absence of atrial stretch . In the fibrotic atria of transforming growth factor-beta (TGF-β) transgenic mice and in dogs with pacing-induced heart failure, slower and more organized atrial tachyarrhythmias were documented [66,67]. In our recent preliminary study,  we induced heart failure by ventricular tachypacing (220bpm) in sheep and investigated the dynamics and mechanisms of AF. Echocardiography confirmed the development of congestive heart failure and a significant increase of the LA size (from 31±1 in controls to 50±1mm in failing hearts) at 6–8 weeks after pacing. Then, we conducted optical mapping in the isolated hearts by recording from the endocardial surface of the PLA, which was exposed by a minimal incision applied to the LA. The DF of excitation during sustained AF episodes induced in the presence of 0 to 4μM acetylcholine was much slower in failing hearts than in controls; importantly, in both cases the DFmax at the PLA was significantly larger than the DFs of bipolar electrograms recorded from the Bachmann\'s bundle (BB) and RAA (Fig. 8A). Fig. 9A shows sequentially obtained phase maps from a representative experiment during an AF episode. Fig. 9B shows a schematic of the breakthrough sites (color dots) in the PLA during AF induced without acetylcholine in failing hearts. In all of the hearts that were examined, the majority of breakthrough sites were located at the periphery of the PLA. In failing hearts, the activation patterns were mainly repetitive breakthroughs appearing at the PV ostium and propagating toward the periphery of the field of view (Fig. 9A); in some cases, long-lasting stationary micro-reentrant waves were also observed, which anchored to the PVs (Fig. 9C and D). The biatrial electrograms obtained by recording differentially from both atrial appendages (LAA and RAA) showed slower and irregular activity, whereas the PV electrograms were monomorphic and highly periodic, suggesting that a stable source existed within the PV ostial region. Histological analysis of the PLA revealed that interstitial fibrosis was significantly greater in failing hearts compared with controls (Fig. 8B). More importantly, the fibrotic areas tended to cluster in relatively large patches in failing hearts, but not in controls. In addition, the fibrotic patches observed in failing hearts were preferentially located in the periphery of the PLA close to the PV ostium. Thus, large patches of fibrosis may have the ability to anchor micro-reentrant sources in failing hearts. Taken together, these preliminary data suggest that the AF dynamics in this heart failure model are governed by the distribution of fibrosis.