Vol. 290, Issue 2, 694-701, August 1999

The Relationship between the Myocardial Kinetics of Meperidine and Its Effect on Myocardial Contractility: Model-Independent Analysis and Optimal Regional Model¹

Richard N. Upton, Yi Fei Huang² , Laurence E. Mather³ and David J. Doolette

Department of Anaesthesia and Intensive Care, Royal Adelaide Hospital, University of Adelaide, Adelaide, Australia (R.N.U., Y.F.H., D.J.D.); and Department of Anaesthesia and Intensive Care, Flinders Medical Center, Flinders University of South Australia, Bedford Park, Australia (L.E.M.)

The myocardial kinetics of meperidine and the relationship between these kinetics and the effect of meperidine on myocardial contractility (maximum positive rate of change of left ventricular pressure) were examined by analysis of previously published data collected in sheep after the i.v. injection of 100 mg of meperidine over 1 s. There was significant hysteresis between reductions in myocardial contractility and the arterial concentrations of meperidine, but not the coronary sinus blood (effluent from the heart) or calculated myocardial concentrations. The peak reduction in contractility occurred after the peak arterial concentration, at the time of the peak myocardial concentration, but before the peak coronary sinus concentration, suggesting that the site of drug action in the heart was not in equilibrium with either arterial blood or effluent blood from the heart. The most appropriate form of a dynamic model (a linear model with a threshold) was determined, without the need to assume a kinetic model, by directly fitting the observed reductions in myocardial contractility to the calculated myocardial concentrations. To determine the optimal kinetic and combined kinetic-dynamic models, a variety of one-, two-, and three-compartment models of the myocardium were fitted to the coronary sinus concentrations by using hybrid modeling. These included "tank in series" models that accounted well for drug dispersion and "peripheral compartment" models that accounted well for deep distribution. The most appropriate model was a "compilation" model, which incorporated features of both these extremes and was a better fit to the observed data than either a traditional single flow-limited compartment or a traditional membrane-limited model.