Thromboxane is produced in response to intracoronary infusions of complement C5a in pigs. Cyclooxygenase blockade does not reduce the myocardial ischemia and leukocyte accumulation.

Activated polymorphonuclear leukocytes (PMNs) contribute to myocardial injury during ischemia and reperfusion. There is evidence that activation of the complement pathway may be one of the mechanisms of PMN activation during ischemia. Intracoronary infusion of complement C5a during normal perfusion pressure is associated with decreased coronary flow, contractile dysfunction, and PMN accumulation. The mechanisms responsible for these changes have not been identified. Thromboxane A2 (TXA2) is a potential mediator of this myocardial ischemic response. Activated PMNs produce TXA2, a known coronary vasoconstrictor, and TXA2 was shown to be a mediator of the pulmonary hypertensive response to activated complement. The goal of the present study was to determine if an enhanced TXA2 production is associated with the myocardial response to C5a and whether cyclooxygenase blockade would reduce the myocardial ischemia. In open-chest pigs, intracoronary C5a (500 ng) caused reversible reductions in blood flow (50.0% of control), regional contractile function (25.8% of control), leukocyte trapping (1.0 x 10(6) cells/g myocardium or a peak artery-coronary venous difference of 5.3 x 10(3) cells/microliters blood), and increased coronary venous TXB2 (the TXA2 breakdown product) from 1.6 pmol/ml to a peak of 6.9 pmol/ml. Cyclooxygenase blockade with aspirin or indomethacin, which prevented TXB2 production, did not alter the response in flow, function, or PMN trapping. Ibuprofen, a known direct inhibitor of PMNs in addition to its cyclooxygenase blockade effect, reduced the response slightly. The pig coronary vascular bed was responsive to the TXA2 agonist U46619, which reduced flow and function without PMN trapping. Mechanical reductions in coronary flow to levels equivalent to those during the C5a infusions did not increase coronary venous TXB2 nor cause PMN trapping but did cause equivalent contractile dysfunction. Incubation of whole blood with C5a at concentrations equivalent to those achieved in vivo did not cause TXB2 production. We conclude that 1) TXA2 is produced in response to intracoronary C5a and 2) cyclooxygenase blockade does not prevent the C5a-induced myocardial ischemia, contractile dysfunction, and PMN trapping. The TXA2 production likely involves a vascular site or a blood cell-vascular interaction. This model system indicates the potential for persistently activated PMNs to cause continued ischemia during myocardial reperfusion.