Time dependency of ventricular fibrillation thresholds determined using trains of stimuli

When determining ventricular fibrillation threshold (VFT), considerable time elapses from beginning a VFT measurement until the actual occurrence of ventricular fibrillaion (VF). We have defined this elapsed time as the "time to VF" and examined the effect of varying time to VF on the measured value of VFT. We induced VF in anesthetized dogs with a 100-Hz train of 16 4-ms stimuli applied to the right ventricular epicardium. The time to VF was varied by changing either the increment of fibrillation current increase (0.5, 1.0, or 2.0 mA) from one train to the next and/or the number of atrial paced beats between trains (6-192 beats). For 0.5-mA current increments, as time to VF increased from 66 to 454 s, VFT fell progressively from 15.3 +/- 4.8 to 6.7 +/- 1.1 mA. When time to VF exceeded 454 s, VFT increased again. Current increments of 1.0 and 2.0 mA had a similar time dependency of VFT. For any time to VF the VFT was lower when small current increments (i.e., more trains) were used to induce VF. Pretreatment with reserpine in six dogs abolished the time dependency. We conclude that time elapsed from the beginning of a VFT determination until VF actually occurs and the total number of trains used are important determinants of the VFT, probably because of local catecholamine release by the stimuli.