Internal cardioversion of atrial fibrillation in sheep.

BACKGROUND The cardioversion efficacy of multiple defibrillation waveforms and electrode systems was compared in a sheep model of atrial fibrillation. METHODS AND RESULTS Sustained atrial fibrillation could be induced with rapid atrial pacing in 23 (55%) of the animals. This study was performed in four parts. Six sheep with sustained atrial fibrillation were used for data analysis for each part, except in part 4 where five sheep without sustained atrial fibrillation were used. In part 1, four lead systems and four single capacitor truncated exponential defibrillation waveforms (two monophasic and two biphasic) were tested. In part 2, two transvenous lead systems were compared; one was a right-to-left system with one electrode located in the right side of the heart and the other electrode located in the left side of the heart, and the other was a totally right-sided system with both electrodes located in the right side of the heart. Eight (four monophasic and four biphasic) waveforms were tested with each lead system. In part 3, eight transvenous lead systems were compared, and two waveforms (one monophasic and one biphasic) were tested with each lead system. For parts 1-3, probability of success curves were determined for each waveform/lead system configuration using an up-down technique with 15 shocks per configuration. In part 4, shocks were synchronized to the QRS and given through two lead configurations during sinus rhythm in 20-V steps starting with 40 and ending with 500 V, and two waveforms were tested with each lead system (one monophasic and one biphasic). Ventricular fibrillation thresholds were determined by giving shocks during the T wave of sinus rhythm. For part 1, the three lead systems that used only intravenous catheter electrodes had significantly lower defibrillation requirements than the catheter-to-chest wall patch system. A 3/3-msec biphasic waveform had significantly lower defibrillation requirements than any of the other three waveforms in part 1. In part 2, the 3/3-msec biphasic waveform with a right-to-left lead system configuration had significantly lower defibrillation requirements than any other waveform lead system combination tested, and for each waveform tested, the right-to-left configuration had significantly lower requirements than the totally right-sided configuration. In part 3, for each waveform the right-to-left configuration had significantly lower voltage and energy requirements than the corresponding totally right-sided configuration. Furthermore, in part 3, waveform/lead configurations that probably generated high potential gradients near the sinoatrial node and near the atrioventricular node resulted in more postshock conduction disturbances. In part 4, there were no episodes of ventricular arrhythmias with shocks synchronized to the QRS. However, with synchronization to the T wave, ventricular fibrillation was induced in all five animals with the minimum tested voltage, which was 40 V. CONCLUSIONS This acute model yielded sustained atrial fibrillation in approximately 55% of the animals. Cardioversion of atrial fibrillation in sheep is possible with very low energy requirements using transvenous electrode systems (50% successful energy of 1.3 +/- 0.4 J for the 3/3-msec biphasic waveform with a right-to-left lead system). The biphasic waveform had the lowest defibrillation requirements of any waveforms tested, and right-to-left lead systems resulted in lower defibrillation requirements than totally right-sided lead systems. Also, lead systems that probably generated high potential gradients near the sinoatrial and atrioventricular node areas resulted in more frequent episodes of postshock conduction disturbances. Furthermore, synchronization of the shock to the QRS was vital to avoid potentially lethal postshock ventricular arrhythmias...