The pathophysiologic sequence leading to respiratory failure after chest tr
auma can be an inevitable consequence of the primary injury or a secondary,
mediator-driven inflammatory process. To distinguish between these alterna
tives, a simple cross-transfusion experiment was performed. A captive bolt
gun injured the chest of anesthetized pigs that were mechanically ventilate
d with Fio(2) =.21, .50, or .50 plus indomethacin (5 mg/kg intravenous; 15
min before injury). Tube thoracostomy immediately followed. After 30 min, b
lood from these injured donors was transfused into three matched groups of
naive recipients (n = 8, 6, and 4, respectively) for a 33% exchange transfu
sion. Two control groups received blood from uninjured donors with tube tho
racostomies only (Fio(2) = .21, n = 7; Fio(2) = .50, n = 10). Within 15-30
min after transfusion, in recipients from injured donors versus controls, l
ung compliance was decreased 20%, stroke volume and cardiac output were dec
reased 50%, and pulmonary vascular resistance was increased >300% (all p <
.05). These changes recovered to baseline within 60-90 min. The stable meta
bolite of thromboxane A2, thromboxane B2, increased >500% in plasma within
15 min and remained elevated for >120 min. All responses were similar at 21
% or 50% O-2 which suggests that hypoxia per se is not a cause of mediator
production. All responses were eliminated by indomethacin. By 24 h, histolo
gic changes included atelectasis in 3/3 recipients from injured donors vers
us 0/3 recipients from uninjured donors. We conclude that 1) blunt chest tr
auma releases blood borne mediators, including prostanoids; 2) these mediat
ors can cause secondary cardiopulmonary changes in naive recipients similar
to those produced by chest trauma; 3) the progression to trauma-induced re
spiratory failure is multifactorial; 4) early pharmacologic intervention, r
ather than supportive care alone, may benefit some victims of severe chest
trauma.