Introduction

Restoring blood flow after an ischaemic period is essential to limit myocardial damage. However, reperfusion can lead to fatal arrhythmias, presumably due in part at least to heterogeneities in electrophysiology (EP) across the affected tissue. The aim of this work was to elucidate mechanisms underlying ischaemia-reperfusion (I/R) arrhythmias.

Methods

Single camera mirror-based panoramic optical mapping of Langendorff-perfused rabbit hearts allowed us to obtain high-speed recordings of epicardial transmembrane voltage across the whole heart by means of a voltage-sensitive dye (di-4-ANBDQPQ) used in conjunction with a heart-surrounding LED illumination (640 nm wavelength). An N₂ bubbled bath kept the hearts at a physiologic temperature while minimizing otherwise possible transmural O₂ diffusion into the tissue from ambient air. The hearts were perfused both globally (via the aorta) and locally (via cannulation of the anterior branch of the left circumflex coronary artery) with an oxygenated physiological saline solution in which local perfusion was switched to and from solutions that mimic ischemia (i.e., hypoxemia + hyperkalaemia + acidosis), individual aspects of ischemia, or to no-flow.

Experimental findings were used to inform a computational model of I/R developed in CARP to explore arrhythmia inducibility. This permitted us to reproduce the aforementioned experimental conditions to investigate re-entry mechanisms.

Results

We observed preferential recovery of electrical excitability along the main branch of the reperfused coronary vessel (‘perivascular excitation tunnelling’, PVET) upon switching the local perfusate from both no-flow or ischemic solution to normal solution. In a subset of hearts, this resulted in re-entrant arrhythmias. Regarding experiments with individual aspects of ischemia, local hyperkalaemia and hypoxemia individually presented a strong substrate for PVET and re-entrant arrhythmia occurrence during I/R, while local acidosis did not lead to PVET events.

The I/R computational model reproduced PVET and illustrated quantitative plausible mechanisms underlying re-entry. In keeping with experimental observations, simulations predicted reperfusion arrhythmias after local ischaemia, hyperkalaemia or hypoxemia. Reperfusion from local acidosis alone did not result in arrhythmogenesis. Computational simulations predict that gradual switching between a solution mimicking ischaemia and standard saline would be less arrhythmogenic than a rapid switch.

Conclusions

We observed a novel PVET-based re-entry mechanism upon coronary reperfusion, suggesting that potentially detrimental reperfusion-induced gradients can arise along acutely reperfused coronary vessels. Gradual recovery of perfusate properties may be less arrhythmogenic than instant recovery of physiological flow.