Coronary steal in awake dogs: a real phenomenon

Abstract

Nine mongrel dogs were instrumented with left circumflex coronary artery (LCf) flow probes and balloon occluders and the LCf narrowed so that peak hyperaemia following release of a 15 s coronary occlusion was diminished by 40.4±5.9%. Following insertion of a 16-gauge catheter through an LCf branch immediately distal to the stenosis and advancement to the branch's junction with the main vessel, the left anterior descending artery (LAD) was ligated in its middle third. The dogs were studied 7 to 14 days following recovery from surgery. Haemodynamics and myocardial blood flow were measured under control conditions and again during adenosine infusion, 280 to 320 μg·min⁻¹, through the LCf catheter. After the start of treadmill running measurements before and during adenosine infusion were repeated. At rest the gradient across the LCf stenosis averaged 2.4 kPa. During adenosine infusion there were no detectable systemic haemodynamic effects, but average distal LCf pressure fell from 11.4 to 7.3 kPa in the eight dogs responding to the vasodilator. This decrease in coronary perfusion pressure was associated with a significant decrease in collateral flow. Under control conditions ischaemic LAD flow averaged 60% of normal flow. Adenosine doubled coronary flow to the LCf areas directly perfused with this agent, while the average flow to the collateral-dependent area beyond the LAD stenosis decreased by 20%. The single dog with no increase in LCf flow or fall in distal LCf pressure induced by adenosine also had no change in collateral flow. During exercise LCf flow again double and distal LCf pressure diminished. Ischaemic LAD flow did not change in 3 of 6 exercising animals despite two- to three-fold increases in flow to other left ventricular areas. The fall in ischaemic LAD flow was obscured in three other animals by changes accompanying the exercise state, eg, increased aortic pressure and increased perfusion of collaterals not affected by the coronary steal occurring in the distal LCf vascular bed. Thus, the arteriolar dilating agent adenosine increases coronary flow and decreases perfusion pressure resulting in a coronary flow redistribution, or coronary steal. The metabolic vasodilatation of exercise also creates the haemodynamic milieu for a coronary steal, but other factors accompanying the exercise state may hide the expected decrease in ischaemic flow.