The impact of altered ventricular activation, including biventricular (BV) pacing, on T-wave alternans (TWA) and arrhythmic substrates is unclear. We studied how differing ventricular activation sequence alters TWA; the interval from peak-to-end of the T-wave (TpTe) and T-wave amplitude during right (RV), left (LV), and biventricular (BV) pacing; and right atrial (RA) pacing in patients with preexisting conduction delay.Methods and Results
We measured TWA during RA, RV, LV, and BV pacing in 33 patients receiving cardiac-resynchronization-therapy-defibrillators. TWA magnitude (Valt) was lower during BV than RV (P < 0.01), RA (P < 0.01), or LV pacing. As a result, BV-TWA was more often negative than RV-TWA (P < 0.01), LV-TWA, and RA-TWA, particularly when discordant between pacing modes (P < 0.01). Overall, 83% of TWA recordings were abnormal (25% indeterminate), and 17% negative. BV pacing reduced T-wave amplitude (P < 0.05) and TpTe (P < 0.005) compared to RV pacing and LV pacing (P < 0.05; P < 0.005 respectively). Notably, TWA magnitude varied linearly with T-wave amplitude for all pacing modes (P < 0.001). Over 410 ± 252 days' follow-up, RV-TWA predicted the combined endpoint of death and ICD therapy with 86% negative predictive value (P < 0.05). BV-TWA, RA-TWA, and other repolarization indices were not predictiveConclusions
BV pacing attenuates TWA in tandem with reduced T-wave magnitude. In these patients with baseline QRS prolongation, RV-TWA predicted events more effectively than BV-TWA and RA-TWA. Further studies are required to understand how altered ventricular activation influences repolarization dynamics and arrhythmic tendency.