Higher-Order Dynamics Beyond Repolarization Alternans in Ex-Vivo Human Ventricles are Independent of the Restitution Properties

Abstract

Background: Repolarization alternans, defined as period-2 oscillation in the repolarization phase of the action potentials, provides a mechanistic link between cellular dynamics and ventricular fibrillation (VF). Theoretically, higher-order periodicities (e.g., periods 4, 6, 8,...) are expected but have minimal experimental evidence.

Methods: We studied explanted human hearts obtained from recipients of heart transplantation at the time of surgery. Optical mapping of the transmembrane potential was performed after staining the hearts with voltage-sensitive fluorescent dyes. Hearts were stimulated at an increasing rate until VF was induced. Signals recorded from the right ventricle endocardial surface prior to induction of VF and in the presence of 1:1 conduction were processed using the Principal Component Analysis and a combinatorial algorithm to detect and quantify higher-order dynamics. Results were correlated to the underlying electrophysiological characteristics as quantified by restitution curves and conduction velocity.

Results: A prominent and statistically significant global 1:4 peak (corresponding to period-4 dynamics) was seen in three of the six studied hearts. Local (pixel-wise) analysis revealed the spatially heterogeneous distribution of periods 4, 6, and 8, with the regional presence of periods greater than two in all the hearts. There was no significant correlation between the underlying restitution properties and the period of each pixel.

Discussion: We present evidence of higher-order periodicities and the co-existence of such regions with stable non-chaotic areas in ex-vivo human hearts. We infer from the independence of the period to the underlying restitution properties that the oscillation of the excitation-contraction coupling and calcium cycling mechanisms is the primary mechanism of higher-order dynamics. These higher-order regions may act as niduses of instability that can degenerate into chaotic fibrillation and may provide targets for substrate-based ablation of VF.

Figure 1.. The schematics of global analysis.

Spatiotemporally processed signals recorded from multiple points on a line, showing staggered action potentials upstrokes consistent with wavefront propagation (A). Same signals as A shifted to align the upstrokes (B). The first principle component $\left(W_1\right)$ displays alternans (C). Spectrogram of C, showing a 1:2 peak of alternans (D). The second principle component $\left(W_2\right)$ displays more pronounced alternans larger than C (E). Spectrogram of E, shows a prominent 1:2 peak of alternans (F). The bar in A depicts 200 ms.

Figure 2.. The schematic of composite restitution curve generation.

An exponential curve fits a set of (DI, APD) data points for one pixel. The 95% confidence-interval curves are also shown (A). The restitution curves for all the pixels in a region of interest (say, with period-2 on the local analysis) are calculated (only 100 curves are shown here)(B). Monte Carlo sampling is performed to generate a composite restitution curve (the thick green line) and the confidence interval (shown in a green shade)(C).

Figure 3.. Comparison of baseline and pre-VF spectrograms using global analysis.

The blue spectrograms are the baseline (stimulation cycle length of 500 ms except for H4 at 800 ms) and the orange spectrograms are obtained just before VF induction. H1, H2 (A and B) exhibit prominent 1:4 peaks (the red arrows), while no discernable 1:4 peak is seen for H6 (D). H4 has a ~0.18 peak (the green arrow), corresponding to mainly period-6 activity (C). The baseline signals are multiplied by 0.1 to offset the signals for better visualization.

Figure 4.. The pacing frequency dependence of 1:2 and 1:4 peaks.

In both panels, the 1:2 peak (the classic APD alternans) starts when the cycle length decreases to ~400 ms. On the other hand, the 1:4 peak only rises above the baseline once the cycle length decreases to 300–350 ms. Also, note the different scaling of the 1:4 peak compared to the 1:2 peak, which is 2–3 orders of magnitude smaller than the 1:2 peak. These results significantly reduce the chance that the observed 1:4 peaks are processing artifacts and point to their dynamical origin.

Figure 5.. Examples of signals with different periodicities.

Representative pixel-level optical-mapping signals with period-2 (A), period-4 (B), period-6 (C), period-8 (D), and higher-order/chaotic (E) are shown. Note the intermittency of period-8.

Figure 6.. The distribution of higher-order areas.

In H1, there are multiple period-4 areas (red) in a sea of classic alternans (blue)(A). H2 shows both 1:4 and 1:6 (green) areas (B). H5 has large areas of period-6 without significant period-4 (C). The main feature of H6 is a large area of period-8 without significant 1:2 or 1:4 regions (D). The background images (gray-colored) represent anatomy. Atr, the atrial; Ven, the ventricles; El, the pacing electrode.

Figure 7.. The relative frequency of different periodicities.

Each histogram shows the relative proportions of pixels with a given period (in the range 1 to 8) for each of the six hearts.

Figure 8.. The electrophysiological characteristics of areas with different periodicities.

The top row (A1, B1, C1) compares restitution curves, whereas the other panels show the electrophysiological properties (restitution curve parameters and conduction velocity) between different periodicities. Each panel applies to one heart. Only one comparison (B2) shows any significant difference in the baseline electrophysiological properties, with a marginal separation in the corresponding restitution curves (B1).