Length of excitation wave and susceptibility to reentrant atrial arrhythmias in normal conscious dogs.

We calculated the wavelength of the atrial impulse in chronically instrumented conscious dogs by measuring both conduction velocity and refractory period: wavelength = refractory period X conduction velocity. Implantation of multiple stimulating and recording electrodes allowed wavelength determination at four different areas: the right and left parts of Bachmann's bundle and the free walls of the right and left atria. During programmed electrical stimulation, three types of arrhythmias were observed: rapid repetitive responses, atrial flutter, and atrial fibrillation. During normal rhythm, the wavelength of the atrial impulse varied between 14 and 18 cm. Premature beats had a shorter wavelength, depending on the degree of prematurity. Premature beats that evoked rapid repetitive responses showed a critical shortening of the wavelength below 12.3 cm. Episodes of atrial flutter were induced at a wavelength below 9.7 cm, while fibrillation occurred at wavelengths shorter than 7.8 cm. We correlated the induction of these arrhythmias with the values of refractory period, conduction velocity, and wavelength during control and during administration of several drugs. Intravenous administration of acetylcholine shortened the wavelength by 30-40%, mainly because of refractory period shortening. Both propafenone and lidocaine had strong but opposite effects on refractoriness and conduction and, consequently, little effect on the wavelength. Quinidine markedly prolonged the refractory period, but prolongation of wavelength was less because of a simultaneous decrease in conduction velocity. d-Sotalol also increased refractory period, but because it had no appreciable effect on conduction velocity, this drug was the most effective in prolongation of wavelength. Linear discriminant analysis of the data showed that the refractory period and the conduction velocity each were poor parameters to predict the occurrence of the different arrhythmias (predictive value 48% and 38%, respectively). The combination of both properties, however, as expressed in the wavelength, was a more reliable index that predicted the induction of the different arrhythmias correctly in 75% of the cases. We conclude that the wavelength is a useful parameter for evaluating antiarrhythmic drugs.