Mechanisms of arrhythmogenesis related to calcium-driven alternans in a model of human atrial fibrillation

Figure 1. Human atria model. (a) Single cell APD restitution for ALTfast (solid line) and ALTslow (dashed line) AP models, using left atrium (LA, green) or right atrium (RA, orange) chronic AF K+ current densities from Grandi et al.45 Bifurcation of the APD restitution curve indicates alternans occurring at that pacing CL. (b) LA cell APs for the ALTfast and ALTslow models at different pacing CLs. (c) RA cell APs for the ALTfast and ALTslow models at different pacing CLs. (d) LA and RA regions in the human atria model. (e) Stimulus locations. (f) Detection of filaments during arrhythmia.

Figure 2. Discordant APD alternans during S1 pacing and induction of reentry by an S2 stimulus. (a) APD alternans magnitude. Regions of opposite phase are shown in different colors (red vs. blue). (b) Time to 90% repolarization after even (left) and odd (right) paced beats. (c) Activation time maps after an S2 stimulus was delivered from RS0 (see Fig. 1E) at a coupling interval of 280 ms following an even S1 beat. (d) Transmembrane potential (Vm) maps showing conduction block and reentry following the S2 stimulus in panel (c). Top rows: ALTfast. Bottom rows: ALTslow.

Figure 3. Vulnerability of atria to reentry. Summary of results for ALTfast (a) and ALTslow (b) simulations. The duration of post-S2 arrhythmia was quantified by the number of times that reentry propagated around the pulmonary veins (referred to as a reentrant cycle). The period of reentries that sustained beyond one reentrant cycle was approximately 250 ms. When post-S2 reentrant activity lasted longer than three reentrant cycles, arrhythmia broke up into multiple wavelets (labeled “3+”). The simulation indicated by the white cross had arrhythmia lasting longer than 2 s.

Figure 4. Discordant APD alternans and filaments during fast pacing. (a) APD alternans magnitude in ALTfast (top) and ALTslow (bottom) during pacing at 270-ms CL. Regions of opposite phase are shown in different colors (red vs. blue). (b) Nodal surfaces (black) and filaments (color) in ALTslow during 260-ms CL pacing. Snapshots show filament locations at different times after a paced beat from SAN region. Filaments from the first four beats are shown in different colors on the same snapshot for comparison.

Figure 5. Filament dynamics during fast pacing. (a) Heatmaps of the number of distinct filaments present over time in ALTfast (left) and ALTslow (right). (b,d,e): Histograms and adjusted box-whisker plots for ALTfast (red) and ALTslow (blue). Green crosses indicate outliers. (b) Number of filaments during arrhythmia. Fraction of time indicates the number of time points during which a particular number of filaments was present, normalized by the total number of time points during which any number of filaments greater than zero was present. (c) Cumulative number of new filaments formed in ALTfast (solid line) and ALTslow (dotted line) at pacing CLs of 270 ms (red), 260 ms (blue), and 250 ms (green). (d) Filament duration. (e) Filament tree (FT) duration.

Figure 6. Composition of FTs during fast pacing. (a) Graphs of FT branching structure from two example FTs in ALTfast (top) and ALTslow (bottom) simulations. Filament duration is indicated by node colors; time is not drawn to scale. (b) Trajectories of ALTfast (left) and ALTslow (right) FTs from panel (a). (c) Correlation between FT duration and FT composition metrics. Top: ALTfast. Bottom: ALTslow. Left to right: mean filament duration, number of filaments, maximum filament duration, and sum of filament durations. Linear best fit lines and Pearson correlation coefficients are shown in each plot.