Abundant evidence has accumulated during recent years to indicate that intravascular coagulation may be the determining event in the pathogenesis of the generalized Shwartzman reaction.

In the presence of reticuloendothelial blockade, the intravenous injection of a protein antigen into specifically immunized rabbits or the infusion of soluble immune complexes into normal animals has been shown to result in the production of bilateral renal cortical necrosis. The similarity in the pathogenesis of this lesion and that seen in the classical generalized Shwartzman reaction produced by bacterial endotoxins is indicated by (a) the failure of both lesions to develop in animals pretreated with large doses of heparin, (b) by the finding of "heparin-precipitable fibrinogen" in the circulation, and (c) by the presence of massive fibrin deposits within the glomerular capillaries. These findings indicate that antigen-antibody reactions in vivo are capable of activating the blood coagulation system and that the mode of action of bacterial endotoxins may have an immunological basis.

PLATES 83 AND 84(Received for publication, July 6, 1962) When thrombin is administered intravenously to rabbits for several hours at a rate just insufficient to produce intravascular thrombosis, there is nevertheless conversion of much of the animal's circulating fibrinogen to fibrin (1). In most instances, no detrimental effect is observed, and no extensive fibrin deposition is found (2). This absence of pathologic alterations in the face of extensive intravascular fibrin formation has generally been ascribed to the efficient lysis of fibrin by plasmin and other blood proteolytic enzymes. While local activation of plasminogen may be responsible in the rabbit for eliminating fibrin already deposited, recent evidence (2) has indicated that the bulk of fibrin aggregates formed in the circulating blood are rapidly removed by the reticuloendothelial system in a manner similar to the efficient clearance of foreign particulate matter and colloidal materials. It was shown that blockade of the reticuloendothelial system in rabbits receiving intravenous thrombin infusions regularly resulted in massive fibrin deposition in the terminal vascular bed. The major lesion, bilateral renal cortical necrosis, was indistinguishable from that of the generalized Shwartzman reaction which has classically been elicited by bacterial endotoxins.In the present report immunohistochemical studies have provided direct evidence for the phagocytic clearance of fibrin by demonstrating the accumulation of this material selectively within the reticuloendothelial cells of the liver and spleen after the administration of thrombin or endotoxin. These observations further substantiate the hypothesis (2) that the generalized Shwartzman reaction in the rabbit results from the failure of the reticuloendothelial system to adequately clear the circulation of fibrin aggregates formed during systemic blood coagulation initiated by endotoxin.

No abstract

Intradermal injections of bacterial endotoxins in normal rabbits ordinarily yield delayed inflammatory reactions which are not manifest in the gross for several hours and reach maximal development only after a day or more. While the mechanism underlying these characteristic cutaneous reactions is not understood, it has been pointed out that the reactions resemble, in many respects, those of delayed hypersensitivity (1). The present study was prompted by recent chance observations which suggested that the dermal reactivity of rabbits to bacterial endotoxins could be much modified by previous exposure of the animals to endotoxins. It will be shown that prior injection of endotoxins leads to the development of Arthus-type reactivity and that this is correlated with the development of an immune response of unexpected qualitative and quantitative characteristics. The present report is concerned with a general description of the accelerated cutaneous reactivity, its passive transfer with serum and its possible significance. Materials and MethodsAnimals.--Albino hybrid rabbits of either sex weighing approximately 1.6 to 1.8 kilos were used. Commercial rabbit pellets and water were allowed ad libitum.Endotox/ns.--Lyophilized lipopolysaccharide endotoxins from several Gram-negative bacteria were obtained from the Difco Laboratories, Inc., Detroit, and were dissolved in physiologic saline solution for injection. Volumes of 2.0 ml. for intravenous injections and 0.2 ml. for intradermal injections were used throughout. Intradermal injections were made in the upper abdominal quadrants, from which hair had been removed with dectfic dippers.Histdogic Studies.--Samples of skin to be examined histologically were removed immediately following sacrifice of the anima.ls, fixed in Bouin's solution and stained with hematoxylineosin. Serologic Stud/es.--Qualitative precipitin reactions were carried out in capillary precipitin tubes, using bacterial endotoxins in concentrations of 10 to 50 mg./ml, mixed with equal