Short-term action potential memory and electrical restitution: A cellular computational study on the stability of cardiac repolarization under dynamic pacing.

Massimiliano Zaniboni,Elena Tolkacheva

doi:10.1371/journal.pone.0193416

Abstract

Electrical restitution (ER) is a major determinant of repolarization stability and, under fast pacing rate, it reveals memory properties of the cardiac action potential (AP), whose dynamics have never been fully elucidated, nor their ionic mechanisms. Previous studies have looked at ER mainly in terms of changes in AP duration (APD) when the preceding diastolic interval (DI) changes and described dynamic conditions where this relationship shows hysteresis which, in turn, has been proposed as a marker of short-term AP memory and repolarization stability. By means of numerical simulations of a non-propagated human ventricular AP, we show here that measuring ER as APD versus the preceding cycle length (CL) provides additional information on repolarization dynamics which is not contained in the companion formulation. We focus particularly on fast pacing rate conditions with a beat-to-beat variable CL, where memory properties emerge from APD vs CL and not from APD vs DI and should thus be stored in APD and not in DI. We provide an ion-currents characterization of such conditions under periodic and random CL variability, and show that the memory stored in APD plays a stabilizing role on AP repolarization under pacing rate perturbations. The gating kinetics of L-type calcium current seems to be the main determinant of this safety mechanism. We also show that, at fast pacing rate and under otherwise identical pacing conditions, a periodically beat-to-beat changing CL is more effective than a random one in stabilizing repolarization. In summary, we propose a novel view of short-term AP memory, differentially stored between systole and diastole, which opens a number of methodological and theoretical implications for the understanding of arrhythmia development.

Highlights

At a given steady-state, if cycle length (CL) suddenly changes to another constant value, adaptation of its duration (APD) will instantly change according to ERCL and, after a given number of beats (Nb), reach a new steady-state given by RDCL; in other words, it takes a different Nb for each Electrical restitution (ER) curve to rotate into the rate dependence (RD) curve
On we will call the average value of conditioning CL, “average value of conditioning pacing CL (CLÃ)”, and define the state of the membrane for a given nth beat with one point in the (CL, APD)- or (DI, APD)-space with coordinates (CLn-1, APDn) or (DIn-1, APDn); all the information needed to assign the state of the n+1th beat is in the ERCL obtained after the nth beat
The link between hysteresis and repolarization stability has been previously described in the case of dERDI [25], our results show that dERCL can be used to understand and predict, for instance, conditions when otherwise silent pathologies of ion currents might become proarrhythmic, and how pacing rate variability can control the transition to arrhythmias

Summary

Introduction

Several mechanisms contribute to adjusting the partition of the cycle length into diastole and systole, the mechanical counterparts of the ventricular AP and DI, for the cardiac pump to combine pacing rate with strength of contraction and meet the organism’s energy requirements. It is the repolarization phase of AP that controls the relaxation of the heart chambers, and the kinetics of ion currents involved in this phase determine the rate-dependent adaptation of its duration (APD) [1]. HRV is considered a risk marker in several cardiac pathological states [9], the interest in characterizing the complex rate-dependent behavior of ventricular APD in non-stationary pacing conditions by looking at dynamic restitution properties, rather than stationary ones

Methods

Results

Discussion

Conclusion