Studies on the Mechanisms Involved in the Production of Abortion by Endotoxin

Open Access
Publisher Website
Google Scholar
Abstract
Discussion and Summary: Escherichia coli lipopolysaccharide in doses as low as 5 μg was shown to produce abortion in 85% of pregnant mice. This phenomenon was found to be accompanied by hemorrhage into the decidua basalis and a striking degree of congestion of the placental labyrinth. When mice deliver live young spontaneously, this congestion of the placental spaces is not seen. The aborting effect of endotoxin probably does not occur simply as a result of making the mother quite ill. Zahl and Bjerknes (4), using a Shigella extract, were able to abort pregnant mice without any associated mortality. The present study has demonstrated that severe shock, in the form of glycogen anaphylaxis, does not induce abortion. In addition, it has been shown that the lethal effect of endotoxin and the aborting effect of endotoxin can be dissociated. Cortisone protects against the lethal effect, but not against the abortion reaction. The possibility exists that the primary site of the aborting action of endotoxin is on the fetus. It may be that endotoxin crosses the placental barrier, kills the fetus and this results in abortion. However, this explanation appears unlikely in view of the well known resistance of young animals to the lethal effects of endotoxin (1, 3, 18). According to Zahl and Bjerknes (4), histologic studies on the fetuses of endotoxin treated mice revealed no pathologic alterations. Neither were lesions found in the maternal ovary or corpus luteum. Failure of various pharmacologic agents to prevent endotoxin-induced abortion is submitted as evidence that the phenomenon dealt with here is not analogous to the local and generalized Shwartzman reactions or the epinephrine skin lesion of Thomas. One possible prototype for the abortion reaction was originally suggested by Zahl and Bjerknes (4). They were impressed by the similarity between endotoxin-induced abortion and endotoxin-induced hemorrhage in malignant tumors. They remarked that intravenously injected colloidal dyes localized in both the tumors and in placentas. In addition, the time interval between endotoxin administration and hemorrhage is approximately the same for both reactions. Algire et al. (19) studied endotoxin-induced tumor hemorrhage by use of a transparent chamber technique. They remarked that tumor necrosis was preceded by slowing of the circulation with stasis and venous engorgement. A similar reaction may occur in the placenta as indicated by the present histologic studies. Certainly, engorgement was a prominent characteristic of the endotoxin treated placentas. There may very well be a common underlying physiologic basis behind the action of endotoxin on these two types of invasive foreign tissue. Finally, this study has shown that the abortion reaction is not a placental analogue of either the local or generalized Shwartzman reactions or the epinephrine skin lesions. In addition, evidence has been presented which shows that abortion does not depend on the lethal shocking effect of endotoxin. Indeed, it appears that the abortion reaction is a specific type of endotoxin-induced lesion, mediated by an as yet undefined mechanism, and displaying certain features in common with endotoxin-induced tumor hemorrhage.

This publication has 7 references indexed in Scilit: