PERIPHERAL BYPASS-INDUCED PULMONARY AND CORONARY VASCULAR INJURY - ASSOCIATION WITH INCREASED LEVELS OF TUMOR-NECROSIS-FACTOR

Background. Although cardiopulmonary bypass is associated with systemi c complement activation and neutrophil sequestration, it is unclear wh ether bypass-induced vascular injury is localized and dependent on org an ischemia. We hypothesized that other factors perhaps related to pla cement of a bypass circuit or to blood perfusion of a pump-oxygenator system may produce vascular injury caused by systemically circulating mediators. In dogs, we determined whether application of a systemic ve noarterial bypass circuit with pump-oxygenator perfusion but without p ulmonary or cardiac flow diversion (peripheral bypass) leads to vascul ar injury. Since several features of the postperfusion syndrome after bypass resemble sequelae of endotoxin exposure, we also measured circu lating endotoxin and tumor necrosis factor levels. Methods and Results . Anesthetized dogs underwent 2 hours of exposure to a pump-oxygenator with peripheral venoarterial bypass. We used a double indicator measu rement of pulmonary and coronary vascular permeability (protein leak i ndex [PLI]) as indexes of vascular injury. Compared with controls (n=7 ), the pulmonary PLI of dogs undergoing bypass (n=11) increased more t han threefold (18.8 +/- 2.3 vs 63.3 +/- 7.6 x 10(-4) min-1; P<.05) and the coronary PLI increased more than twofold (P<.05). The rate of dis appearance of intravascular radiolabeled protein increased threefold a fter bypass (disappearance t1/2, 241 +/- 35 vs 84 +/- 15 minutes, cont rol vs bypass; P<.05), suggesting a generalized increase in vascular p ermeability. Circulating endotoxin was detectable in blood samples fro m 8 of 8 bypass animals (range, 0.24 to 4.56 ng/mL) compared with 2 of 5 controls (P<.05). Tumor necrosis factor levels increased significan tly with bypass (6.7 +/- 3.8 vs 146.7 +/- 33.6 U/mL, baseline vs bypas s; P<.05) and were only slightly and nonsignificantly increased in con trols (7.0 +/- 4.4 vs 18.2 +/- 5.9 U/mL; P=NS). Peak tumor necrosis fa ctor but not peak endotoxin levels correlated with pulmonary and with coronary protein leak. As expected, circulating complement (CH50) leve ls decreased significantly during bypass, reflecting systemic compleme nt activation. However, the levels correlated poorly with the severity of vascular injury. Conclusions. We conclude that peripheral pump-oxy genator bypass causes coronary and pulmonary vascular injury that is i ndependent of blood flow diversion and is associated with the appearan ce of circulating levels of endotoxin and tumor necrosis factor, which may play a role in bypass-induced vascular injury.