Thus, one or more streptococcal superantigens may contribute significantly to TSLS. The pathogenesis of other severe invasive streptococcal diseases remains obscure. To date, few studies have evaluated transmission of the TSLS organism, rates of carriage and infection, and serosusceptibility to SPEs. Such studies are greatly needed. Recently, we received 23 strains of group B streptococci from patients with TSLS, 4 group C strains, 2 group F strains, and 9 group G strains. In these patients, S. aureus and group A streptococci were not isolated despite extensive searching; thus, their treating physicians proposed that the illnesses were caused by non-group A strains. Also, many isolates were obtained from otherwise sterile body sites, increasing the likelihood of their involvement in disease. Thus far, we have purified one pyrogenic toxin superantigen from a group B streptococcal strain and nearly purified one from a group F strain. There also appears to be a significant increase in the number of highly severe and fatal infections in bone marrow transplant patients, with symptoms resembling TSLS and with a-hemolytic streptococci being isolated. These strains should be examined further for production of pyrogenic toxin superantigens. In this regard we have partially purified one pyrogenic toxin from a Streptococcus sanguis isolate. Several mechanisms by which the pyrogenic toxin superantigens cause the most severe symptoms ofTSS, notably the hypotension and shock, have been proposed. It is generally thought that these illnesses are capillary leak syndromes. In rabbit models for the same illness, very small amounts of fluid provided protection from challenge with lethal amounts of the purified toxins. Although capillary leak explains the hypotension, there is considerable debate concerning the mechanism ofinduction of that capillary leak. The four current hypotheses are superantigenicity with consequent release of lymphokines and monokines, monokine release alone, direct capillary leak, and enhancement of endotoxin shock. Superantigenicity with consequent release oflymphokines and monokines has been most studied.
Cytokines such as TNFa are thought to be actively involved in endotoxin shock and thus may contribute significantly in TSS illnesses also. Fast et al. have shown that pyrogenic toxins induce a sustained release of TNF, and this effect depends on the presence of T cells. Subsequently, Lee et al. showed that T cells do not have to proliferate during TSS and TSLS induction in experimental animals, suggesting that TNF release from macrophages does not depend on either T cell proliferation or cytokine release despite a T cell requirement. This is based on the ability of cyclosporin A to block T cell proliferation but not to block the lethality of the toxin in animals. Lee et al. also provided evidence in vitro that TSST-1 can alter capillary permeability directly. Finally, all pyrogenic toxins amplify the lethal effect of endotoxin by as much as 105 – to 106-fold.
Since humans have a predominantly gramnegative intestinal flora and there appears to be an increase in vaginal gram-negative organisms in TSS patients, it is possible that endotoxin enhancement contributes to the hypotension or at least significantly increases the severity of illness. As little as 2 ILg of endotoxin in humans can cause symptoms of endotoxin shock. Thus, if that effect were potentiated by 106-fold, only picogram amounts of endotoxin would be necessary to cause similar symptoms. Leonard and Schlievert have recently shown that SPE A binds very strongly to endotoxin, and the complex, once formed, is lethal to immune cells. This effect may contribute to the nearly complete disappearance ofimmune cells in patients with the most severe forms of TSS and TSLS.