A Novel Approach to Measure Short Term Cardiac Ventricular Action Potential Memory: Comparison between Five Numerical Models

Abstract

Rate-dependence (RD) and electrical restitution (ER) describe the dependence of cardiac action potential duration (APD) from the average or from the previous pacing cycle length (CL). We know however that in certain dynamic conditions APD depends not only from the average of from the previous, but from a given small number (3 to 5) of CLs preceding each beat. Such dependence is usually called short term AP memory (STAM) and has been shown to control repolarization stability. Despite the relevance of STAM for the transition to arrhythmias, a clear definition of it and a way to measure it are still lacking, making it difficult to compare findings from different preparations/laboratories. By means of simulations carried out on 5 different human cardiac ventricular AP models, I provide a novel general definition of STAM which is substantially different from those proposed in literature so far, and is based on the collective behavior of the family of ER curves that describe the system during dynamic pacing. I show how STAM can predict repolarization stability under high- and beat-to-beat-variable pacing rate after singular perturbations of the rhythm. The measure of STAM in the 5 AP models shows that memory involves a number of preceding beats up to 15 at high pacing rate (CL = 320 ms). I monitored STAM over consecutive beats during periodic and random pacing trains in a compact representation that allows comparison between models. STAM, as measured here, compactly represents otherwise hidden dynamic properties of cardiac AP at high pacing rate, which is of great interest in a variety of cardiac electrophysiological issues, like the mechanism underlying latent instability of repolarization in LQT1 syndrome, to cite one.