Perivascular Excitation Tunnelling: a Novel and Preventable Cause of Cardiac Reperfusion Arrhythmias

Background: Reperfusion of ischaemic myocardium can lead to ischaemia-reperfusion arrhythmias (IRA). The mechanisms underlying IRA are incompletely understood.

Methods: Isolated hearts (n = 97 rabbits; n = 16 pigs) were loaded with a voltage-sensitive dye for optical mapping of sub-epicardial action potential dynamics. Hearts were simultaneously perfused ‘globally’ (Langendoff perfusion via the aorta) and ‘locally’ (via cannulation of a single coronary artery), initially with oxygenated physiological saline. Local perfusion was then either stopped (no-flow ischaemia [NF]), or switched to solutions mimicking various aspects of acute myocardial ischaemia (hyperkalaemia [HiK+], hypoxia [LoO2], acidosis [HiH+], or a combination of the three factors for simulated ischaemia [SI]) to give rise to an electrically inexcitable tissue volume. Recovery of excitability upon reperfusion, and emergence of IRA were optically tracked. A modified two-step reperfusion strategy to reduce IRA incidence was established in rabbits and then validated in pig hearts to assess the potential clinical relevance of observed IRA mechanisms and potential treatment strategies.

Results: Local NF gave rise to formation of an inexcitable ischaemic tissue volume within ~45 min. This could be mimicked by local perfusion with SI solution (block within ~2-5 min), or individually by HiK+, LoO2, but not HiH+. Upon reperfusion of the locally ischaemic tissue, we observed preferential recovery of electrical excitability in myocardium along the reperfused vessel’s main branch (‘perivascular excitation tunnelling’, PVET) in ~86 of cases (NF: 83%; SI: 89.7%; HiK+: 80 %; LoO2: 40 % in rabbits and HiK+: 89% in pigs). In about half of these cases, PVET resulted in re-entry.

PVET-based re-entry was prevented by a two-step reperfusion approach, where first of the distal part of the previously inexcitable tissue was re-exposed to physiological saline until excitable (~10 min), before the entire previously ischaemic volume was reperfused from the proximal end of the cannulated coronary artery. With this strategy, PVET in distal tissue did not give rise to re-entry, as it was blocked by the still inexcitable proximal tissue, whereas upon reperfusion of the proximal tissue, the path length for PVET was too short to sustain re-entry.

Conclusions: We identify PVET as a novel mechanism underlying IRA upon coronary reperfusion, and show that two-step reperfusion can prevent PVET-induced re-entry. Future work will be dedicated to progressing from two-stage (sequential) reperfusion to two-zone (simultaneous) reperfusion of the affected vascular bed, where distal exposure to physiological solution is combined with proximal perfusion of oxygenated cardioplegic solution.