Temporal characteristics of cardiac memory in humans: vectorcardiographic quantification in a model of cardiac pacing.

OBJECTIVES: The purpose of this study was to assess the temporal characteristics of cardiac memory in a human pacing model.
BACKGROUND: Cardiac memory is induced by periods of altered ventricular activation and in the canine pacing model develops in 2 to 3 weeks.
METHODS: Cardiac memory development (phase 1) and resolution (phase 2) was followed qualitatively (ECG) and quantitatively (vectorcardiography [VCG]) in 20 patients with symptomatic sinus bradycardia receiving DDD-R pacing at physiologic rates. During phase 1, maximum ventricular pacing in the right ventricular apical endocardium was achieved by short AV delay. ECG and VCG were recorded during normal ventricular activation prior to implantation (sinus rhythm), 1 day after and then weekly for 5 to 8 weeks during AAI pacing (n = 14, "long-term"), and daily during 1 week (n = 6, "short-term"). In phase 2, the remaining cardiac memory was related to the amount of ventricular pacing. A long AV delay was chosen to reduce ventricular pacing, and 14 patients were seen once after 4 to 5 weeks.
RESULTS: ECG and VCG showed marked changes in all patients within 1 week's pacing and remained stable during phase 1. Of the 14 long-term patients, 13 showed significant change in T vector azimuth (mean -150 degrees ) and 11 in T vector elevation (mean 63 degrees ). At the end of phase 2, a linear relation existed between remaining cardiac memory (%) and amount (%) of delivered ventricular pacing.
CONCLUSIONS: Cardiac memory developed and reached steady state within 1 week of right ventricular endocardial pacing at physiologic rates in man, was preserved in proportion to delivered ventricular pacing, and completely resolved within 1 month in its absence.