In vivo assessment of left atrial contractile performance in normal and pathological conditions using a time-varying elastance model.

Abstract

BACKGROUND Contractile function of the ex vivo, isolated left atrium (LA) has been described by a time-varying elastance, but this atrial chamber property has not been shown in vivo. METHODS AND RESULTS Instantaneous LA pressure-volume (P-V) relations were studied in 12 anesthetized, autonomically blocked, atrially paced dogs. LA volume was calculated from orthogonal sonomicrometer pairs using a cast-validated formula. Data were collected during increases in LA pressure produced by a phenylephrine bolus (200 to 400 micrograms IV). Isochronal P-V points from 5 beats, representing a wide range of atrial pressures, were fitted by linear regression analysis (range of R2, .92 to .99). There were significant time-dependent increases in the slopes [E(t)] and small but statistically insignificant decreases in the volume axis intercepts [VO(t)] of the instantaneous LA P-V relations during atrial contraction; maximal elastance (Emax) occurred 29 +/- 16 milliseconds before atrial end systole (minimal LA volume). Emax was not significantly different than the slopes of either the nonisochronal end-systolic P-V relation (Ees) or the nonisochronal maximal P-to-V relation (EmaxPV): 5.5 +/- 2.8, 4.3 +/- 1.5, and 5.4 +/- 4.2 mm Hg/mL, respectively. In 7 dogs, data were collected both before and after a rapid infusion of calcium gluconate (1 to 2 g IV). Emax increased significantly with a calcium-induced increase in inotropic state (4.5 +/- 1.6 to 5.7 +/- 1.8 mm Hg/mL, P < .01), but the volume axis intercept was unchanged (3.6 +/- 0.7 versus 3.4 +/- 1.9, P = NS). In 4 additional dogs with heart failure (mean LA pressure, 26 +/- 6 mm Hg) produced by 3 weeks of rapid right ventricular pacing, LA stroke volume was significantly greater than and elastance determinations were similar to those of normal dogs. However, the effects of calcium infusion on LA function were attenuated in these animals. CONCLUSIONS We conclude that (1) in the intact heart, LA contraction may be approximated by time-varying elastance with time-dependent changes in E(t) and that (2) LA systolic P-V relations using either the nonisochronal maximum P-to-V ratio or end systole may be useful as an estimate of Emax, are highly linear and sensitive to calcium-induced changes in inotropic state, and may be useful in identifying LA chamber adaptation to chronic hemodynamic loads.