Influence of Holter monitor and electrophysiologic study methods and efficacy criteria on the outcome of patients with ventricular tachycardia and ventricular fibrillation in the ESVEM trial

Becausenot all laboratories use the monitoring and stimulation protocols used in the Electrophysiologic Study Versus Electrocardiographic Monitoring (ESVEM) trial, we reanalyzed the ESVEM patients' data using alternative, commonly used Holter monitor (HM) and programmed stimulation efficacy criteria to determine if different criteria would have changed the trial's conclusions. Also, because β-blocker use and coronary artery disease frequency were not equally distributed between the two limbs in ESVEM, we reanalyzed the ESVEM data adjusting for the possible effect of these variables. In the HM limb, drug efficacy in the original ESVEM analysis was declared by reduction of total premature ventricular complexes (PVCs) by 70%, pairs by 80%, runs of 3 to 15 beats by 90%, and all ventricular tachycardia (VT) more than 15 beats by 100%. In this analysis, we examine outcome in subjects meeting two more stringent sets of criteria, (1) reduction of total PVCs by 70%, of pairs by 80%, and of all VT by 100% (new criteria set 1) and (2) reduction of total PVCs by 80%, of pairs by 90%, and of all VT by 100% (new criteria set 2). In electrophysiology (EPS) limb patients, we compared arrhythmia recurrence when efficacy was declared with triple extrastimuli as compared with maximally testing with double extrastimuli, and arrhythmia recurrence was compared in patients tested with identical versus any more aggressive protocol on drug than was used before drug. We also compared the predictive accuracy of zero versus 3 to 15, and 0 to 5, 6 to 10, and more than 10 induced beats on drug. Additionally, we compared predictive accuracy of the HM- and EP-guided limbs excluding patients on β blockers and those with noncoronary disease. Lastly, to determine whether concordant results on HM and EPS testing would provide more accurate efficacy predictions than EP testing alone, HM recordings obtained in EPS-limb patients but not processed or used during the course of the ESVEM study were analyzed. The original ESVEM HM criteria, new set 1, and new set 2 yielded predicted drug efficacy rates of 77%, 68%, and 58%, respectively; however, arrhythmia recurrence rates were unchanged. Similarly, arrhythmia recurrence rates for patients tested with triple versus less than triple extrastimuli ( P = .238), more aggressive versus identical protocols ( P = .955), and 0 to 5 v 6 to 10 v more than 10 induced beats ( P = .263) or 0 v 3 to 15 induced beats ( P = .106) were unchanged. In the 215 (of 286) patients with coronary disease and not receiving β blockers, there was still no difference in arrhythmia recurrence or mortality between the noninvasive and invasive limbs in ESVEM. Lastly, in patients with drug efficacy predictions by EPS testing, there was no difference in outcome in patients who had concordant versus discordant efficacy predictions by simultaneously obtained HMs. The use of more stringent testing methods and efficacy criteria would not have significantly improved the predictive accuracy of drug assessment by HM or EPS in the ESVEM trial. Additionally, excess noncoronary disease in EP-guided patients and excess β-blocker use in HM-guided patients did not influence the results in the ESVEM trial.