48 patients (40 Male), mean age 68 +/- 8 years, in III-IV class, with intraventricular conduction delay, received a biventricular pacemaker. Heart failure aetiology was non-ischemic in 60%. Left ventricular lead positioning was inferior in 5 patients (10%), posterior in 12 (25%), lateral in 18 (37%) and anterior in 13 (27%). QRS duration and axis were evaluated in sinus rhythm, and during right ventricular pacing, left ventricular pacing and biventricular pacing, the last early after implant and late after 8.8 +/- 4.3 months. QRS duration (ms) was 154 +/- 29 in sinus rhythm, 175 +/- 28 during right ventricular pacing, 196 +/- 31 during left ventricular pacing, 122 +/- 23 during biventricular pacing "early" and 120 +/- 18 during biventricular pacing "late." All the differences were statistically significant, but not between "early" and "late" biventricular pacing. Mean QRS axis ( degrees ) was -27 +/- 32 in sinus rhythm, -75 +/- 4 during right ventricular pacing, 112 +/- 41 during left ventricular pacing, -82 +/- 51 during biventricular pacing "early" and -80 +/- 42 during biventricular pacing "late." Only the difference between left ventricular pacing and all the other groups was statistically significant. QRS axis did not significantly differ according to left ventricular lead site during left and biventricular pacing. "Late" compared with "early" biventricular pacing axis showed variation >30 degrees in 35% of patients, in spite of no significant changes in QRS duration and x-ray positioning.
Conclusion: Biventricular pacing significantly reduced QRS width, which persisted long-term. Left and biventricular pacing axis was poorly related to left ventricular lead positioning. Biventricular pacing axis variability over time may suggest a role of electrical remodeling.