MICROVASCULAR REACTIVITY AFTER CRYSTALLOID, COLD BLOOD, AND WARM BLOOD CARDIOPLEGIC ARREST

Background. The effects of three techniques of cardioplegic arrest on endothelium-dependent microvascular function of the right and left ventricles were examined in a canine model of cardiopulmonary bypass. Methods. Oxygenated cold crystalloid cardioplegia and cold blood cardioplegia groups, (n = 11 each) had hypothermic cardiopulmonary bypass (28 degrees C), topical cooling, antegrade arrest, and intermittent antegrade delivery. A warm blood cardioplegia group (n = 11) had normothermic cardiopulmonary bypass (37 degrees C), antegrade arrest, and continuous antegrade delivery. All groups underwent cardioplegic arrest for 1 hour followed by 1 hour of reperfusion. Dogs that did not have instrumentation were used as controls (n = 10). Coronary microvessels (100 to 200 mu m in internal diameter) were examined in a pressurized, no-flow state with video microscopic imaging and electronic dimension analysis. Results. Ischemic arrest with cold crystalloid cardioplegia significantly (p < 0.05) impaired endothelium-dependent relaxations in both ventricles to acetylcholine (left ventricle, 69% +/- 4%, and right ventricle, 73% +/- 5%, versus control left ventricle, 100% +/- 0.3%, and control right ventricle, 100% +/- 0.3%) and the calcium ionophore (left ventricle, 70% +/- 6%, and right ventricle, 68% +/- 3%, versus control left ventricle, 98% +/- 1%, and control right ventricle, 98% +/- 1%). In the cold blood cardioplegia group, endothelium-dependent relaxations to acetylcholine (left ventricle, 96% +/- 1%, and right ventricle, 87% +/- 4%) and the calcium ionophore (left ventricle, 88% +/- 3%, and right ventricle, 78% +/- 7%) were preserved. In the warm blood cardioplegia group, endothelium-dependent responses to acetylcholine (92% +/- 3%) and the calcium ionophore (96% +/- 1%) were preserved in the left ventricle, but the right ventricle showed reduced (p < 0.05) reactivity to the endothelium-dependent acetylcholine (77% +/- 8%) and the calcium ionophore (69% +/- 8%). Endothelium-independent relaxation to sodium nitroprusside was similar to controls in all groups for both ventricles, thus indicating normal smooth muscle responsiveness. Conclusions. Cardioplegic arrest with cold blood cardioplegia preserved the endothelium-dependent response in the right and left ventricles, whereas cold crystalloid cardioplegia impairs this response. Warm blood cardioplegia preserved the endothelium-dependent response in the left ventricle, but this response was reduced in the right ventricle. This suggests that blood cardioplegia and hypothermia may be important in protection of microvascular endothelium and that the right ventricle may be more vulnerable to damage than the left ventricle.