1. Cortisone, in a dose of 25 mg. daily and with a pretreatment period of 3 days, in rabbits weighing 1 to 1.5 kilos, did not inhibit the dermal Shwartzman reaction produced by meningococcal or S. marcescens toxin.
2. In cortisone-treated rabbits, a single intradermal injection of toxin produced a primary reaction of hemorrhage and necrosis in the skin at the injected site. This lesion resembled the Shwartzman reaction in its gross and histological appearance.
3. Like the Shwartzman reaction, the primary hemorrhagic reaction in cortisone-treated rabbits was prevented by nitrogen mustard, and the preventive effect of nitrogen mustard was partly eliminated when the femoral marrow was protected against the latter agent.
4. A single intravenous injection of meningococcal or S. marcescens toxin, in cortisone-treated rabbits, was followed by bilateral cortical necrosis of the kidneys in the majority of instances. The renal lesions, as well as hemorrhages in the lungs, spleen, liver, and gastrointestinal tract, resembled the lesions of the generalized Shwartzman reaction. Histologically, the glomerular capillaries in both types appeared to be occluded by homogeneous, eosinophilic material which showed a strongly positive Schiff reaction.
5. The renal lesion following a single injection of toxin in cortisone-treated animals, and that following two intravenous injections in the generalized Shwartzman reaction, were both completely prevented by nitrogen mustard. This effect of nitrogen mustard was inhibited when the femoral marrow was protected against the latter agent.
6. The injection of S. marcescens toxin into the skin of normal rabbits did not cause systemic symptoms, nor was it possible to provoke the generalized Shwartzman reaction by this route. In cortisone-treated rabbits, a similar intradermal injection was regularly followed by the development of bilateral cortical necrosis of the kidneys, indicating that absorption of toxin from the skin occurred in these animals.
7. Possible mechanisms to account for the observations are discussed.
The authors are obliged to Professor James R. Dawson for many helpful suggestions during the course of this investigation.