Stochasticity in Action Potential duration Enhances Dispersion of Repolarisation at Fast Pacing Rates

Patch clamp recordings of isolated ventricular myocytes frequently display a temporal variability in the action potential duration. Intercellular coupling through gap junctions has been shown to reduce such variability which may return as coupling is reduced. A phenomenological model capable of accurately capturing action potential morphology and variability observed in experimental data has been developed with tunable restitution properties. The phenomenological formulation also allows fast simulation compared with biophysically detailed models. This model is now used to investigate the effects of action potential variability over prolonged periods. A two dimensional tissue slab is simulated using a monodomain model and the simulation software Chaste. Simulations are performed with and without variability in the action potential model. Different coupling strengths are used with a physiological conductivity corresponding to a conduction velocity of 71cm/s. Simulations are performed at full conductivity, and with the conductivity scaled by factors of 0.5, 0.1 and 0.05. At a pacing cycle length of 1000ms with physiological coupling and differences in action potential duration with and without variability are negligible. Under reduced coupling, the difference increases to a maximum of 2ms. No correlation is observed between beats. At a cycle length of 230ms, temporal variability drives the cell model to alternans. This effect is reduced in tissue at all conductivities. Tissue was paced for 40 beats at a cycle length of 230ms before a cycle length reduction of 4ms. The process was repeated until propagation failed. At a cycle length of 222ms a spatially discordant alternans of magnitude 2ms was observed. In conclusion, stochastic effects are masked at physiological conductivity values and effects do not accumulate over time. At significantly reduced conductivity heterogeneities in repolarisation can be induced