We developed a canine model for the in vivo utilization of angiogenesis factors to promote revascularization of a collateral-dependent area of the heart and assessed the potential of heparin in this preparation. Ameroids were placed on the proximal left anterior descending coronary artery (LAD) of 29 dogs, and the left internal mammary artery (IMA) was implanted in an intramyocardial tunnel in proximity to the LAD. A tube positioned in the distal IMA provided a continuous retrograde infusion directly into the vessel from an implanted pump. Heparin (15 or 150 U/h) or saline vehicle was infused. After 8 wk, regional myocardial blood flow was assessed in the anesthetized state during adenosine-induced vasodilatation, before and during occlusion of the IMA. The IMA provided a greater proportion of maximal collateral flow in heparin-treated dogs (22 +/- 5%, n = 17) than in saline-treated dogs (9 +/- 2%, n = 12, P less than 0.05). Thus continuous infusion of heparin promotes the formation of collaterals between the extracardiac artery and the myocardial circulation, establishing the feasibility of targeting angiogenic agents for myocardial revascularization.

The purpose of this investigation was to determine whether blood vessels could develop de novo between an extracardiac artery and a collateral-dependent zone of the heart and to quantify the nutritive blood flow afforded by the new vessels. We also adapted the preparation so that angiogenically active agents could be chronically administered directly to the site of neovascularization in subsequent studies. To induce neovascularization between a systemic artery and the coronary circulation, the left internal mammary artery (IMA) was implanted in an intramyocardial tunnel in proximity to the left anterior descending coronary artery (LAD). A tube situated in the distal IMA connected to an implanted pump provided for continuous intra-arterial infusion at the site of angiogenesis. During the same procedure, an ameroid constrictor was placed on the proximal LAD, rendering its perfusion territory collateral dependent during a 2-3 week period. After 8 weeks, the functional capacity of the anastomoses established between the implanted IMA and the LAD territory was assessed by determining regional myocardial blood flow under basal conditions, during adenosine-induced vasodilatation, and during differential occlusions of the IMA and left circumflex coronary artery (LCCA). For all dogs, IMA occlusion decreased maximal LAD territory flow from 1.31+0.11 to 1.16±0.10 ml/min/g (p

Intramyocardial implantation of a systemic artery [the internal mammary artery (IMA)] causes angiogenesis, with formation of systemic to coronary anastomoses. In dogs, we assessed the magnitude of IMA-derived nutritive flow and determined its influence on regional contraction. We also sought to determine whether acidic fibroblast growth factor (FGF), an angiogenic peptide, could enhance myocardial neovascularization. Ameroid constrictors and hydraulic balloon occluders were placed on the left anterior descending coronary artery (LAD) of 23 dogs, and the left IMA was implanted in the LAD territory. Dogs were randomized to receive continuous infusions of acidic FGF with heparin, heparin alone, or placebo directly into the IMA for 8 wk. Regional myocardial blood flow was assessed in the conscious state 3 days and 8 wk after operation. Left ventricular function was determined in the anesthetized state at the 8-wk conclusion of treatment. In all dogs, IMA occlusion reduced mean maximal LAD zone perfusion by 28% (P < 0.001), without influencing regional contraction. When IMA occlusion was superimposed on left circumflex coronary artery (LCX) occlusion, LAD zone perfusion declined by 34% (relative to LCX occlusion alone), significantly impairing regional contraction. Treatment with either acidic FGF plus heparin or heparin alone improved IMA-derived collateral flow; however, addition of acidic FGF to heparin afforded no additional advantage over heparin by itself. We conclude that acidic FGF did not enhance myocardial angiogenesis in this model. IMA-derived collateral flow has significant functional importance; however, it is evident in the dog only when other sources of collateral flow are compromised.