Abstract
Instability of the cardiac electrical system is a major cause of life-threatening arrhythmias. Current safety guidelines have traditionally emphasized QT interval prolongation on the electrocardiogram (ECG) as a primary biomarker of proarrhythmic risk. However, this approach does not capture all forms of electrical instability. In this study, we demonstrate that beat-to-beat variability in action potential duration (APD), a cellular surrogate of the QT interval, better stratifies proarrhythmic risk than mean APD prolongation alone. Using a validated computational model of a human ventricular myocyte, we simulated the effects of the β-adrenergic agonist isoproterenol (ISO). Increasing the ISO concentration shortened mean APD but paradoxically increased beat-to-beat APD variability, measured as the standard deviation of APDs. In our simulations, the emergence of early afterdepolarizations (EADs), abnormal oscillations of the membrane potential during the plateau phase of the action potential, at higher ISO concentrations was consistently preceded by a marked rise in APD variability, even while mean APD was reduced. Mechanistic analysis revealed that ISO-induced enhancement of L-type Ca2+ current (ICaL) and slow delayed rectifier K+ current (IKs) steepens the APD restitution slope, which reflects action potential instability, even when the net effect is a reduction in APD. These results support APD variability as a more sensitive and mechanistically plausible marker of proarrhythmic risk than APD prolongation alone. Our findings suggest that QT interval variability may improve risk stratification and provide a more robust mechanistic foundation for drug safety evaluation.