Arrhythmogenesis in the heart: Multiscale modeling of the effects of defibrillation shocks and the role of electrophysiological heterogeneity

Abstract

The mechanisms of initiation of ventricular arrhythmias as well as those behind the complex spatiotemporal wave dynamics and its filament organization during ventricular fibrillation (VF) are the topic of intense research and debate. Mechanistic inquiry into the various mechanisms that lead to arrhythmia initiation and VF maintenance is hampered by the inability of current experimental techniques to resolve, with sufficient accuracy, electrical behavior confined to the depth of the ventricles. The objective of this article is to demonstrate that realistic 3D simulations of electrical activity in the heart are capable of bringing a new level of understanding of the mechanisms that underlie arrhythmia initiation and subsequent organization. The article does this by presenting the results of two multiscale simulation studies of ventricular electrical behavior. The first study aims to uncover the mechanisms responsible for rendering the ventricles vulnerable to electric shocks during a specific interval of time, the vulnerable window. The second study focuses on elucidating the role of electrophysiological heterogeneity, and specifically, differences in action potential duration in various ventricular structures, in VF organization. Both studies share common multiscale modeling approaches and analysis, including characterization of scroll-wave filament dynamics.

FIG. 1

(Color) (A) Rabbit ventricular model: geometry and fiber orientation (denoted by short white lines). (B) Action potential from the LR1(RV) and LR1(LV) ionic models. (C) The division of the ventricle into an LV region (region with short action potential duration, colored in blue) and RV/septal region (long action potential region, colored in green) for the study of VF maintenance mechanisms. E1 denotes the pacing site from which the ventricles were paced at a 250 ms basic cycle length. E2 is the electrode site used in VF induction, from which 20 Hz frequency pulses were given. (D) Experimental preparation for shock-induced arrhythmia study. The transmembrane potential distribution in the ventricles corresponds to the seventh pacing stimulus (given from the apex).

FIG. 2

(Color) Transmembrane potential distributions at time of shock delivery (panels A and B, anterior epicardial view) and at shock end (panels C and D, anterior epicardial and transmural views) for a 17 V/cm shock applied at coupling intervals (CIs) of 100, 130, 160, and 190 ms for RV⁻ and LV⁻ polarities, respectively. In panels (E) and (F), the epicardial surface has been rendered semitransparent to allow visualization, in pink, of the filaments at the end of the same shock as in panels (C) and (D). Anterior epicardial and apical views are shown in panels (E) and (F). Color scale is saturated, i.e., transmembrane potentials above +20 mV and below −90 mV appear red and blue, respectively.

FIG. 3

Percentage of myocardial nodes that are of transmembrane potential above +20 mV (black bars) and below −90 mV (gray bars) at the end of a 17 V/cm shock applied at CIs in the range 100 to 190 ms.

FIG. 4

(Color) Evolution of postshock electrical activity following an RV⁻ shock of 17 V/cm strength applied at CIs of 100, 130, 160, and 190 ms. Color scale as in Fig. 2.

FIG. 5

(Color) Evolution of postshock electrical activity following an LV⁻ shock of 17 V/cm strength applied at CIs of 100, 130, 160, and 190 ms. Color scale as in Fig. 2.

FIG. 6

(Color) Transmembrane potential distributions on the anterior and the posterior of the ventricles 195 ms after the end of 17 V/cm shocks applied at CI=130 ms (A and C) and CI=160 ms (B and D) for RV⁻ and LV⁻ polarity. Phase singularities on the epicardium are marked with solid black circles. White arrows indicate direction of propagation. Color scale as in Fig. 2.

FIG. 7

(Color) Transmembrane potential distribution for 3 s of VF (in intervals of 1 s) for heterogeneous APD, LV APD, and RV APD ventricular models. The time t=0 corresponds to the end of the VF-inducing high-frequency pacing. For each case, the top image is anterior epicardial view and the bottom image is apical view with the epicardium rendered semi-transparent to show the scroll-wave filaments (colored pink). The dashed black line at t=1000 ms over the heterogeneous APD model transmembarne potential map denotes the border between regions characterized with a different APD. Epicardial phase singularities are marked with solid black circles.

FIG. 8

Average number of filaments (over the 3 s of VF) for heterogeneous APD, LV APD, and RV APD ventricular models. The total average number of filaments is broken down into filaments in the LV, septal, and RV regions. The boundaries of the three different ventricular regions, as used for filament sorting, are outlined below the graph.

FIG. 9

(Color) Top panels: Posterior epicardial views of transmembrane potential distribution in the heterogeneous APD ventricles. Epicardial phase singularities are marked with solid black circles. Direction of wave front propagation is indicated with white arrows. In the top left image, the black arrow points towards the rotor that is the source of the wave fronts propagating towards the RV. Bottom panels: the same view but with the epicardium rendered semitransparent to display scroll-wave filaments (colored pink). In the bottom left image, the black arrow points towards a new filament formation. The same filament (colored in black) is shown in the other two bottom panels; its ends are colored in white to underscore the fact that it is a U-type filament (both ends of the filament are attached to the epicardial surface). The white dashed line at the t=1001 ms and t=1028 ms panels is the border between regions characterized with a different APD.

FIG. 10

(Color) (A) Cross sections through the ventricle with mapped dominant frequency (DF) distribution for each ventricular model. (B) DF ranges, in 1 Hz intervals, and the percentage of the ventricular volume that exhibits a particular DF for each ventricular model. (C) DF ranges, in 1 Hz intervals, and the percentage of ventricular volume in the heterogeneous APD ventricular model, either of long APD (RV/septal region) or of short APD (LV region), that exhibits the particular DF.