Differential effects of arginine vasopressin on isolated guinea pig heart function during perfusion at constant flow and constant pressure.

8-Arginine vasopressin (AVP) is a powerful coronary vasoconstrictor as well as peripheral vasoconstrictor, but AVP also is reported to have negative cardiac inotropic and chronotropic effects in vitro and in vivo. Our aim was to examine the direct effects of coronary vasoconstriction by AVP on cardiac function and metabolism in isolated guinea pig hearts perfused either at a constant perfusion pressure (CPP) of 55 mm Hg or at a constant coronary flow (CCF) equal to the initial natural flow at constant pressure. Coronary vasoconstriction was elicited by perfusing hearts with increasing concentrations of AVP in random order. Variables assessed were atrial heart rate (HR), atrioventricular (AV) conduction time, left ventricular pressure (LVP), coronary flow, inflow and outflow O2 tensions, O2 delivery (Do2), oxygen consumption (MVo2), percentage oxygen extraction (%O2E) and cardiac efficiency (HR-LVP/MVo2). We found that AVP increased coronary vascular resistance more at CCF than at CPP. The decrease in coronary flow, as a function of AVP at CPP, produced concentration-dependent decreases in heart rate, LVP, and MVo2, a decrease in Do2/MVo2, increases in AV conduction time and %O2E, and no significant change in cardiac efficiency. In contrast, the increase in perfusion pressure as a function of AVP at CCF caused no change in HR and AV conduction time, much smaller decreases in LVP and Do2/MVo2, a smaller increase in %O2E, an increase rather than a decrease in MVo2, and a decrease in cardiac efficiency. Our results indicate that larger decreases in HR, LVP, MVo2, and Do2/MVo2, and the larger increases in AV conduction time and %O2E with the AVP-induced decrease in CF at CPP are consistent with myocardial depression resulting from reduced global perfusion. However, cardiac efficiency was maintained at CPP because the decreased HR and LVP product (cardiac work) matched the decrease in MVo2. At CCF, AVP did not directly produce myocardial depression, but the small time-dependent decrease in LVP over time was not matched by the increase in MVo2, so that cardiac efficiency was not maintained. The demonstration of an increase in MVo2 despite no change or a decrease in cardiac work by coronary vasoconstriction with AVP at CCF, but not at CPP, suggests that cardiac O2 use is dependent more on maintenance of CF, despite increased resistance to perfusion, rather than on maintenance of perfusion pressure. Our data agree that Gregg's phenomenon results from a hydraulic effect to distend coronary vasculature because when flow is not allowed to decrease during vasoconstriction, MVo2 increases even though HR is unchanged and LVP is slightly decreased. This is supported by the finding that AVP does not increase coronary vascular resistance during CCF as much as during CPP, so that O2 supply is better maintained to match MVo2.