Ventricular proarrhythmic effects of ventricular cycle length and shock strength in a sheep model of transvenous atrial defibrillation.

BACKGROUND

Synchronized cardioversion is generally accepted as safe for the treatment of ventricular tachycardia and atrial fibrillation when shocks are synchronized to the R wave and delivered transthoracically. However, others have shown that during attempted transvenous cardioversion of rapid ventricular tachycardia, ventricular fibrillation (VF) may be induced. It was our objective to evaluate conditions (short and irregular cycle lengths [CL]) under which VF might be induced during synchronized electrical conversion of atrial fibrillation with transvenous electrodes.

METHODS AND RESULTS

In 16 sheep (weight, 62 +/- 7.8 kg), atrial defibrillation thresholds (ADFT) were determined for a 3-ms/3-ms biphasic shock delivered between two catheters each having 6-cm coil electrodes, one in the great cardiac vein under the left atrial appendage and one in the right atrial appendage along the anterolateral atrioventricular groove. A hexapolar mapping catheter was positioned in the right ventricular apex for shock synchronization. In 8 sheep (group A), a shock intensity 20 V less than the ADFT was used for testing, and in the remaining 8 sheep (group B), a shock intensity of twice ADFT was used. With a modified extrastimulus technique, a basic train of eight stimuli alone (part 1) and with single (part 2) and double (part 3) extrastimuli were applied to right ventricular plunge electrodes. Atrial defibrillation shocks were delivered synchronized to the last depolarization. In part 4, shocks were delivered during atrial fibrillation. The preceding CL was evaluated over a range of 150 to 1000 milliseconds. Shocks were also delayed 2, 20, 50, and 100 milliseconds after the last depolarization from the stimulus (parts 1 through 3) or intrinsic depolarization (part 4). The mean ADFT for group A was 127 +/- 48 V, 0.71 +/- 0.60 J and for group B, 136 +/- 37 V, 0.79 +/- 0.42 J (NS, P > .15). Of 1870 shocks delivered, 11 episodes of VF were induced. Group A had no episodes of VF in part 1, two episodes of VF in part 2 (CL, 240 and 230 milliseconds with 2-millisecond delay), and one episode each in parts 3 (CL, 280 milliseconds with 2-millisecond delay) and 4 (CL, 240 milliseconds with 100-millisecond delay). Group B had two episodes in part 1 (CL, 250 and 300 milliseconds with 20-millisecond delay), three episodes in part 2 (CL, 230, 230, and 250 milliseconds with 2-millisecond delay), and one episode each in parts 3 (CL, 260 milliseconds with 2-millisecond delay) and 4 (198 milliseconds with 100-millisecond delay). No episodes of VF were induced for shocks delivered after a CL > 300 milliseconds.

CONCLUSIONS

Synchronized transvenous atrial defibrillation shocks delivered on beats with a short preceding ventricular cycle length (< 300 milliseconds) are associated with a significantly increased risk of initiation of VF. To decrease the risk of ventricular proarrhythmia, short CLs should be avoided.