Lung injury and neutrophil density during air embolization in sheep after leukocyte depletion with nitrogen mustard

The hypothesis that neutrophils are a major mediator of acute lung injury is based in part on the results from leukocyte depletion studies in which peripheral leukopenia is induced by administration of chemotherapeutic agents or antileukocyte serum. Because of valid concerns about the specificity and multiplicity of effects these regimens have in vivo, we designed a correlative structure-function study of the impact of one depletion regimen on the intrapulmonary density of neutrophils in two groups of sheep. We gave one group of four awake sheep repeated injections of nitrogen mustard (0.4 mg/kg body weight on the ninth, sixth, and third day before the experiment) to lower circulating leukocyte counts to less than 1% of their normopenic value. The other group of four awake sheep was not treated before the experiment (normopenic sheep). On the day of the experiment, the sheep were anesthetized to measure hemodynamics and lung lymph dynamics, and to take lung tissue for morphologic examination before and during 4 h of continuous venous air embolization. Reduction of circulating leukocytes by greater than 99% of normopenic values only lowered pulmonary intravascular neutrophil density by about 80%. In baseline lung biopsies, neutrophil density in small pulmonary arteries (1 to 0.1 mm in diameter) in the nitrogen-mustard-treated sheep averaged 51 +/- 23 (+/- 1 SD) cells/mm2 of blood area versus 291 +/- 40 in the normopenic sheep. During venous air embolization, neutrophils sequestered in the pulmonary arterial microvessels of both groups (210 +/- 27 cells/mm2 of blood area in the nitrogen-mustard-treated sheep versus 1,217 +/- 49 in the normopenic sheep). Parallel increases in neutrophil density occurred in alveolar capillaries and small pulmonary veins. Neutrophils attached to the intravascular air emboli and to microvascular endothelial cells. Endothelial cell gaps were seen in about 30% (13 gaps in 48 microvessels) of the small arteries in the nitrogen mustard-treated sheep versus about 80% (38 gaps in 48 microvessels) in the normopenic sheep. Lung lymph protein clearance nearly doubled in the nitrogen mustard-treated sheep and tripled in the normopenic sheep during venous air embolization compared with the respective baseline clearance values. We suggest that methods used to deplete circulating leukocytes do not assure removal of neutrophils from the lungs; the remaining neutrophils appear capable of responding to intravascular inflammatory stimuli.