Upper limit of vulnerability reliably predicts the defibrillation threshold in humans.

BACKGROUND The upper limit of vulnerability is the stimulus strength above which electrical stimulation cannot induce ventricular fibrillation even when the stimulus occurs during the vulnerable period of the cardiac cycle. The purpose of this study was to test the hypothesis that the upper limit of vulnerability can accurately predict the defibrillation threshold in patients undergoing implantable cardioverter-defibrillator (ICD) implantation using nonthoracotomy lead systems. METHODS AND RESULTS We studied 77 patients at the time of ICD implantation. Multiple endocardial-endocardial and endocardial-subcutaneous shock pathways were used. Two different protocols were used to test the upper limit of vulnerability. In protocol 1 (n = 17), the upper limit of vulnerability was tested with two shocks on the peak or the up-slope of the T wave of paced rhythm. The shocks were given randomly either at the peak and 20 milliseconds before the peak of T wave (n = 7) or at 20 and 40 milliseconds before the peak of T wave (n = 10). In protocol 2 (n = 60), the upper limit of vulnerability was tested with three shocks delivered at 0, 20, and 40 milliseconds before the peak of the T wave. The weakest shock that failed to induce ventricular fibrillation by a 5-J step-down or step-up method was defined as the upper limit of vulnerability. The defibrillation threshold was also determined by a 5-J step-down or step-up method. In protocol 1, the upper limit of vulnerability (9 +/- 6 J) was significantly lower than the defibrillation threshold (13 +/- 7 J) with a correlation coefficient of .87 and P < .001. In protocol 2, the upper limit of vulnerability (13 +/- 6 J) was not significantly different from the defibrillation threshold (13 +/- 6 J) with a correlation coefficient of .85 and P < .001. In 45 of the 60 patients, the upper limit of vulnerability was < or = 15 J; all had a defibrillation threshold of < or = 20 J. In 51 of the 60 patients, the upper limit of vulnerability was within 5 J of the defibrillation threshold. The upper limit of vulnerability overestimated the defibrillation threshold by > 10 J in 8 patients and underestimated the defibrillation threshold by > 10 J in only 1 patient. The overestimation and underestimation occurred only in patients with the upper limit of vulnerability > 15 J. CONCLUSIONS When tested with three shocks on and before the peak of the T wave, the upper limit of vulnerability accurately predicted the defibrillation threshold in patients undergoing ICD implantation using nonthoracotomy lead systems. This method required either one or no episodes of ventricular fibrillation in most patients.