Unexpected action of platelet activating factor antagonism after fluid resuscitation from traumatic shock

Background. We have shown that therapy directed at polymorphonuclear neutrophil (PMN) CD18 receptors attenuates sequelae associated with a post- trauma endotoxin (lipopolysaccharide [LPS]) challenge. Platelet activating factor (PAF) stimulates PMNs by CD18-independent pathways, and WEB-2086, a PAF receptor antagonist, blunts septic symptoms in many experimental models. This study tested the hypothesis that the blockade of non-CD18 dependent PMN adherence attenuates trauma- and LPS-evoked pulmonary, dysfunction. Methods. We performed three experiments. First, anesthetized swine were subjected to hind-limb trauma and 30% hemorrhage. After 1 hour animals were resuscitated with shed blood, lactated Ringer's solution (LRS), and WEB-2086 (10 mg/kg/hr) or vehicle. After a 72-hour recovery period, LPS was administered. LPS was then administered without an earlier episode of traumatic shock to animals treated with WEB-2086 or vehicle. Finally, PAF was infused before and after trauma and a dose response curve was obtained. Results. Surprisingly, PAF blockade increased mortality after trauma (5 of 11 WEB-2086 animals versus I of 9 vehicle animals; p = 0.15) and depressed cardiac index and O₂ delivery at 72 hours (p < 0.05). After LPS administration WEB-2086 treated pigs were unable to manifest the hyperdynamic circulatory compensation seen in the vehicle pigs. In the absence of traumatic shock, WEB-2086 was associated with reduced mortality (four of five WEB-2086 treated pigs versus two of five vehicle pigs survived 5 hours; p = 0.07) and improved arterial Po₂ (p = 0.05) and base excess (p = 0.04) 60 minutes after LPS administration. The dose response curve for PAF infusion on the cardiac index was altered after trauma compared with the nontraumatized state. Conclusions. Because WEB-2086 had unexpected and fundamentally opposite properties before and after trauma, PAF may have a previously undescribed homeostatic role in the compensatory response to injury. These results also suggest that blockade of endogenous inflammatory mediators can have a positive or negative action, depending on the timing of administration and the preexisting condition.