Summary
This study was undertaken to test the hypothesis that granulocytes are needed for
intravascular clotting induced by endotoxin. Rabbits were treated with cortisone in
a dosage that prepares for clotting after a single injection of endotoxin. They were
divided into an experimental group, made granulocytopenic with nitrogen mustard before
the injection of endotoxin, and 3 control groups. One control group was injected with
saline; a second group was injected with endotoxin; a third group was made hypo coagulable
with warfarin before the injection of endotoxin. The controls receiving endotoxin
alone exhibited evidence of slow intravascular clotting lasting several hours: excess
intravascular disappearance of 131I-fibrinogen and falls in fibrinogen levels, platelet counts, factor V and factor
VIII activities. Factor VII and factor XII activities also fell in this group. Fibrin
was seen in glomerular capillaries in many animals. The granulocytopenic animals experienced
neither excess disappearance of intravascular fibrinogen nor fibrin deposition in
the kidneys. Platelet counts fell slowly and clotting factor activities only slightly.
Less evidence of intravascular clotting after endotoxin was found in the granulocytopenic
animals than in the animals made hypocoagulable with warfarin. These data establish
that granulocytes are required for endotoxin to induce significant intravascular clotting
in animals prepared with cortisone.