Functional significance of microvascular collateral anastomoses after chronic coronary artery occlusion.

Abstract A portion of canine left ventricular free wall was rendered collateral dependent by gradual occulsion of the left anterior descending coronary artery (LAD) with a surgically implanted ameroid constrictor. Six to nine weeks after implantation of the constrictor, regional myocardial blood flow was measured with 9-μm microspheres in 12 dogs under control conditions and during diversion of collateral flow retrogradely by venting the LAD to atmospheric pressure. In four dogs, flow measurements during retrograde diversion were also made with 25-μm microspheres. Under control conditions, flow in the collateral-dependent region was similar to that in the normally perfused left ventricular myocardium. Retrograde diversion of collateral flow caused mean aortic pressure to decrease 10%, but did not change heart rate, left atrial blood pressure, or flow in the normally perfused region. Despite diversion of collateral flow from the LAD and the accompanying decrease in aortic pressure, flow in the collateral-dependent region averaged 0.55 ± 0.09 ml/min/g, 67% of the control flow. The 9- and 25-μm microspheres were distributed similarly in transmural sections of the collateral-dependent region, whereas the large microspheres were distributed preferentially toward the endocardium in transmural sections of the normally perfused region. The findings of a large volume of residual, nondivertible collateral flow and of similar transmural distributions of 9- and 25-μm microspheres indicated that a significant portion of the collateral flow traversed microscopic, intramural pathways. A model describing the source of flow to collateral-dependent myocardium was developed. It predicts that approximately one-half of the flow reaches the collateral-dependent region through microvascular pathways.