Mechanistically based mapping of human cardiac fibrillation

Abstract

The mechanisms underpinning human cardiac fibrillation remain elusive. In his 1913 paper 'On dynamic equilibrium in the heart', Mines proposed that an activation wave front could propagate repeatedly in a circle, initiated by a stimulus in the vulnerable period. While the dynamics of activation and recovery are central to cardiac fibrillation, these physiological data are rarely used in clinical mapping. Fibrillation is a rapid irregular rhythm with spatiotemporal disorder resulting from two fundamental mechanisms - sources in preferred cardiac regions or spatially diffuse self-sustaining activity, i.e. with no preferred source. On close inspection, however, this debate may also reflect mapping technique. Fibrillation is initiated from triggers by regional dispersion in repolarization, slow conduction and wavebreak, then sustained by non-uniform interactions of these mechanisms. Notably, optical mapping of action potentials in atrial fibrillation (AF) show spiral wave sources (rotors) in nearly all studies including humans, while most traditional electrogram analyses of AF do not. Techniques may diverge in fibrillation because electrograms summate non-coherent waves within an undefined field whereas optical maps define waves with a visually defined field. Also fibrillation operates at the limits of activation and recovery, which are well represented by action potentials while fibrillatory electrograms poorly represent repolarization. We conclude by suggesting areas for study that may be used, until such time as optical mapping is clinically feasible, to improve mechanistic understanding and therapy of human cardiac fibrillation.

Figure 1. Dynamic balance between activation and recovery in initiating human atrial fibrillation

A, electrograms show ectopic beat (S2) causing AF (varying cycle lengths). B, steep APD restitution curve enables S2 to produce repolarization oscillations preceding AF. C, conduction restitution curve shows dynamic slowing at site of AF onset just prior to AF onset (iii, red slope). D, critical activation delay enables block and formation of a counter‐clockwise spiral initiating AF in right atrium. (From Narayan et al. 2008 b; Schricker et al. 2014.)

Figure 2. Quantitative analysis of electrogram morphology

A, poles of a catheter (bipolar – close, unipolar – distant) record from distinct mapping fields. B, electrograms reflect single wavefront of an organized rhythm (e.g. atrial flutter). C, fibrillation, characterized by an uncertain number of wavefronts of uncertain rate, relative timing (phase) and spatial size in undefined recording fields. Summation of these waves may produce variable electrograms from the same spatiotemporal mechanism, or similar electrograms from variable mechanisms. Accordingly, ‘qS’, ‘rR’, or other electrogram rules in fibrillation are not specific for any particular mechanism. The same argument may apply to unipolar electrograms, which summate across wider regions of tissue.

Figure 3. Limitations of electrogram based activation mapping in AF

A, poles of a clinical bipolar electrode may record unrelated wavefronts in AF. B, fluoroscopic view of co‐localized MAP, bipolar catheters in human atrium. C, MAP in human right atrium indicate local activation (small vertical bars) from far field (asterisks). Notably, bipolar signals (in red) can indicate actual local activation (true positives), show no deflection (false negatives) and show signals that reflect far‐field electrograms (false positives). (Modified from Narayan et al. 2011 b.)

Figure 4. Spiral wave re‐entry as drivers of cardiac fibrillation

A, schematic diagram of spiral wave, showing wavefront curvature as conduction velocity slows towards core (*), where wave front meets wave back. Action potentials from sites 1–3 show varying APD, allowing re‐entry around the unexcited, yet excitable core. (From Pandit & Jalife, 2013.) B, first experimental demonstration of spiral waves in rabbit VF. Phase is depicted in colour with spiral wave chirality indicated by + (clockwise) or – (counter‐clockwise). Three phase singularities (PS) are seen. (From Gray et al. 1998.) C, optical mapping of human atria shows stable micro‐reentrant sources on the endocardium sustaining AF, anchored to endocardial fibre complexity, yet passive, transient activity on the epicardium and elsewhere in the periphery. Optical action potentials (OAPs) at sites 1–4 on the endocardium show activation and recovery over multiple sequential cycles, yet electrograms (Cath 1 EG) vary due to cross‐talk and other factors. The authors concluded that stable endocardial micro‐reentrant sources produce unstable epicardial activations. (From Hansen et al. 2015.)

Figure 5. Human AF maps show differing mechanisms based on mapping technique

A, electrograms from epicardial plaque produce complex maps in human AF. Electrograms used for maps (top) illustrate the challenge of assigning local activity with confidence. No rotational wavefronts were seen in > 4000 maps. (From de Groot et al. 2010.) B, rotor on activation map (early to late) from same group as A, which were sensitive to electrogram type and unstable in this study. (From Lau et al. 2015.) C, focal Impulse and rotor maps (FIRM) showing rotor in human AF, using computational reconstruction of activation and recovery (from physiological MAP and conduction restitution). A stable rotor is identified from phase mapping (here depicted by early meets late activation) that was eliminated by ablation. FIRM‐guided ablation may improve the results of conventional ablation. (From Narayan et al. 2012 c.) D, non‐invasive mapping using the inverse solution show AF sources (progression of phase colours), clustered in stable spatial regions that were treated by localized ablation. Causes for electrical instability at fixed spatial areas may reflect epicardial variability from stable endocardial rotors, technical limitations or other factors. (From Haissaguerre et al. 2014.) E, errors between virtual inverse solution and real contact electrograms in human AF. The first labelled electrogram poorly matched the contact electrogram in timing and shape; the second matched well in timing but not in shape. Timing and shape both influence activation and phase maps. (From Schilling et al. 2000.) See text for further details.

Figure 6. Human VF rotors demonstrated using endocardial and epicardial mapping

A, FIRM mapping using basket electrograms shows human LV rotor 1. B, localized ablation at rotor rendered VF non‐inducible and eliminated VF on long‐term follow‐up. (From Krummen et al. 2015.) C, human VF epicardial rotor (white arrow) on phase map with complex fibrillatory breakdown. (From Nash et al. 2006 b.) D, transmural LV rotor (scroll wave) in ex vivo early human VF displayed using phase maps of endocardial and epicardial electrograms. (From Nair et al. 2011.)