T-wave alternans (TWA) is associated with ventricular tachycardia (VT). Nonalternans repolarization variability (NARV) precedes VT in certain experimental models, but its link to clinical arrhythmia is unproven. This study was conducted to determine if NARV increases prior to VT in patients with implantable cardioverter defibrillators (ICDs).Methods:
TWA and NARV were calculated from shock-channel electrograms preceding onset of VT or non-VT events in patients with an ICD. In each patient, presence of both a VT and a non-VT event with the same QRS morphology before the event was required. Mixed linear model was used for data analysis, using heart rate (HR) and the number of analyzed beats as covariates.Results:
Five hundred and sixty-eight events from 64 patients (males/females 51/13, 67 ± 13 years) were analyzed. HR preceding non-VT events was higher than before VT events (RR interval 595 ± 159 vs 706 ± 111 ms; P < 0.0001). Both TWA and NARV increased with increasing HR (P < 0.001). TWA decreased with increasing number of analyzed beats. When controlled for number of analyzed beats and HR, both TWA and NARV were higher before VT than before non-VT events (P < 0.002 and P < 0.0005, respectively).Conclusions:
NARV is elevated prior to spontaneous VT onset. Both NARV and TWA increase with HR. The decrease of TWA with increasing number of analyzed beats may indicate contamination with NARV or noise when only a small number of beats is available for analysis. NARV might be useful for VT prediction in the future.