Overview	Toxins Involved	Superantigenicity	Symptoms and  Proximal Causes
Case Definitions	Suggested Treatments	Future Treatments	Literature Cited

Toxic-shock syndrome (TSS) is a rapid-onset illness causing fever, hypotension, rash, vomiting, diarrhea, and eventually multiple organ failure.  If not treated promptly, TSS is lethal (Bohach et al. 1990).  TSS is caused by the nonspecific stimulation of T lymphocytes by superantigens that belong to a family of pyrogenic toxins produced by the bacteria Staphylococcus aureus and Streptococcus pyogenes(Schlievert 1993).  Infection with S. aureus produces classical TSS, whereas S. pyogenes causes a modified form of TSS known as either streptococcal TSS (Bohach et al. 1990) or, more recently, toxic-shock-like syndrome (TSLS; Schlievert 1993).  TSLS displays many of the typical TSS symptoms with the addition of severe soft tissue necrosis (Stevens 1995).  For simplicity, staphylococcal TSS henceforth will be referred to simply as TSS, streptococcal TSS will be referred to as TSLS.

Staphylococcal TSS

S. aureusproduces two types of toxins that are implicated in TSS: TSS toxin-1 (TSST-1) and staphylococcal exotoxins (SEs); the latter occur in several serotypes, primarily B and C1.  TSS-related S. aureus infection can be broadly divided into menstrual (those involving vaginal infection in menstruating women) and non-menstrual cases.  Ninety percent of menstrual cases are caused by TSST-1, non-menstrual cases are evenly split between TSST-1 and SEs.  SEs also play a role in staphylococcal food poisoning (Bohach et al. 1990).

Menstrual TSS is usually related to tampon use; tampons increase the vaginal concentration of oxygen, which stimulates TSST-1 production by S. aureus that normally reside in the vagina.  High-absorbancy tampons also sequester magnesium ions, which causes nutrient depletion in the vagina and may simulate late log-phase conditions for resident S. aureus, inducing TSST-1 secretion (Schlievert 1993).  In the late 1970s and early 1980s, a rise in incidence of menstrual TSS was related to use of high-absorbancy tampons, such as the Rely brand, that have since been taken off the market (Schlievert 1993, Bohach et al. 1990).
The dependence of menstrual TSS on TSST-1 could be related to the ability of TSST-1 (but not other pyrogenic toxins) to cross the vaginal mucosa.  Some epithelial cells are known to display TSST-1 receptors; these receptors could be involved in endocytosis, transcytosis, and basolateral secretion of TSST-1 by the vaginal epithelium (Bohach et al. 1990).

Non-menstrual onset of TSS is likely related to opportunistic S. aureus infections in the vagina and elsewhere.  A major class of non-menstrual TSS is flu-associated; S. aureus can infect nasopharyngeal tissues damaged by influenza infection and cause TSS by secreting TSST-1 or SEs (Schlievert 1993).

Streptococcal TSLS

TSLS was first described in 1987 (Bohach et al. 1990), as a result it is not as well characterized as TSS.  Group A streptococcal bacteria, primarily S. pyogenes, cause TSLS by secreting streptococcal pyrogenic exotoxins (SPEs) of three serotypes: A, B, and C.  An additional toxin, streptococcal superantigen (SSA) was recently discovered and may also be involved, though its exact role is unknown.  S. pyogenes infection usually occurs via minor injuries or surgeries, though the exact portal of infection is uknown for most TSLS cases (Stevens 1995).

S. pyogenes occurs in 80 different varieties, many of which produce slightly altered versions of the SPE toxins.  Cross-infection of different strains of S. pyogenes by bacteriophages can lead to a phenomenon similar to antigenic shift in viruses that leads to cyclic outbreaks of highly virulent S. pyogenes.  The emergence of TSLS in the late 1980s may represent such an episode.  In fact, the sudden rise in incidence of soft-tissue necrosis caused by TSLS led to substantial tabloid media coverage of a "flesh-eating bacteria epidemic" in the late 1980s and early 1990s (Stevens 1995).

The various toxins related to TSS and TSLS are structurally homologous to varying degrees.  The SEs are strongly homologous to each other, as are the SPEs, and SEs and SPEs are homologous to each other to a lesser extent (Bohach et al. 1990).  TSST-1 possess the least structural homology to the other toxins -- only 30%.  TSST-1 is also the only pyrogenic toxin that does not require a zinc cofactor (Schmitt et al. 1999).  SE B, SE C, and SPE A share the most structural similarity and are also relatively cross-reactive to the same antibodies.  Interestingly, the three toxins have the most antibody cross-reactivity at their amino termini but share the most structural homology at their carboxy termini.  Several investigations have also produced conflicting data regarding the identity of the terminus responsible for toxic activity (Bohach et al. 1990).  Biochemical data for all the toxins are summarized in Table 1.

All of the TSS- and TSLS-related toxins are able to function as superantigens: proteins that simultaneously bind nonspecifically to T cell receptors (TCRs) and MHC class II molecules outside of the normal peptide-binding groove (Fig. 1; Bohach et al. 1990).  TSST-1, the best characterized of the toxins, contains two structurally distinct binding domains, a b-grasp motif that binds to the V_b CDR2 loop and a b-barrel oligosaccharide/oligonucleotide fold that binds to the MHC class II molecule (Schmitt et al. 1999), possibly using hydrogen bonds (Kum et al. 2000).  Each superantigen exhibits relatively broad specificity for MHC class II molecules, but can only bind to a few (1-5) V_b allotypes (Schlievert 1993).

Figure 1.  Comparison of antigen and superantigen binding to TCR : MHC class II complex.  Left side: normal antigen presentation.  Right side: superantigen stimulation in absence of antigen recognition.  Red = MHC, blue = TCR, green = antigen, purple = superantigen. Adapted from Janeway et al. (1999) and Schlievert (1993).

Superantigen binding of TCRs and MHC class IIs on antigen presenting cells (APCs) activates the T lymphocyte.  Because superantigens bind nonspecifically, polyclonal populations of CD4 T cells are activated and begin proliferation (Schlievert 1993).  Such nonspecific T cell proliferation may be induced by as little as 100 pg/ml of superantigen (Bohach et al.1990).

The large numbers of effector CD4 T cells resulting from this nonspecific proliferation begin stimulating monocytes to secrete several cytokines, including tumor necrosis factor (TNF) a and interleukin (IL) 1.  The nonspecific, high volume superantigen stimulation of T cells results in systemic secretion of these cytokines (instead of the localized secretion that normally occurs during infection), which causes much of the morbidity associated with TSS and TSLS (Janeway et al. 1999, Schlievert 1993).

Some data suggest alternate pathways for TSST-1 stimulation of T cell proliferation.  When APCs were incubated with TSST-1 for 24 h, then washed extensively and introduced to T cells, the APCs were still able to induce T cell proliferation.  This result suggests some sort of internalization and MHC-related presentation of the superantigen, rather than the traditional extracellular superantigen mechanism.  The results of another investigation indicate that TSST-1 activates the IP₃ pathway in a manner similar to that of lectin mitogens (Bohach et al. 1990).  Additionally, TSST-1 may be able to stimulate monocyte secretion of cytokines in the absence of T cell proliferation.  When T cell proliferation was inhibited with cyclosporin, TSST-1 still induced TNF-a production in a lymphocyte/monocyte culture (Schlievert 1993).

The primary symptoms of TSS and TSLS are listed below.  Both syndromes have exceptionally short incubation times, and most of these symptoms will manifest within a matter of 8-12 hours after infection (Chesney 1989).  Effects are summarized in Fig. 2.

Fever.  TNF-a and IL-1 secreted by monocytes act on the hypothalamus to raise the bodily thermostat and cause fever.  TSST-1 may also cross the blood-brain barrier and interact with the hypothalamus directly to induce fever (Bohach et al. 1990).

Susceptibility to Endotoxins.  The pyrogenic toxins increase the host's susceptibility to the endotoxins of other bacteria 10⁵-10⁶-fold.  In the presence of TSS and TSLS exotoxins, only a few picograms of endotoxin are necessary to induce further shock.  Such low levels of endotoxin may be introduced by Gram-negative bacteria already present in the vagina and intestine.  Additionally, SPE A binds to endotoxins, forming a complex that is lethal to immune cells (Schlievert 1993).

Hypotension.  Severe systemic hypotension is caused by a reduction in vasomotor tone, which allows blood to pool in the peripheral tissues, and increased capillary leakage.  Fluid leakage out of the vasculature decreases blood volume and lowers blood pressure.  Leakage is though to be caused one of three mechanisms (or a combination thereof): 1) TNF-a and IL-1 promote vascular permeability, allowing fluid to leak into tissue interstices.  2) TSST-1 alters capillary permeability directly.  3)  Endotoxins (with amplified lethality due to action of pyrogenic toxins) cause capillary leakage directly or indirectly (Schlievert 1993).  Forni et al. (1995) suggest that in TSLS, most hypotension is actually a result of reduced cardiac output rather than changes in the systemic vasculature.

Diarrhea.  Severe diarrhea appears to be mediated directly by pyrogenic toxins (Chesney 1989).

Multiple Organ System Failure.  Reduced tissue perfusion caused by systemic hypotension causes problems for many organs, including the brain, where cerebral edema leads to headaches, disorientation, and confusion.  In a clinical setting, reduced blood volume and dehydration from diarrhea and vomiting require massive fluid replacement that eventually damages the kidneys, causing renal failure, and in some cases respiratory distress.  Impaired circulation to the heart will also cause myocardial failure and death (Chesney 1989).

Erythroderma (rash).  Vascular dilation leads to redness of the epithelia, especially the conjunctiva and sclera of the eye, which may hemorrhage (Chesney 1989).  Pyrogenic toxins are also able to induce degranulation of mast cells and upregulate IgE production (Schlievert 1993); such effects would magnify type I hypersensitivity (Janeway et al. 1999).  TSS also produces long term dermal effects, including desquamation (skin peeling) on the hands and feet after 10-21 days and hair and nail loss after 4-6 weeks (Chesney 1989).

Humoral Immunosuppression.  Despite the size and apparent antigenicity of TSST-1, many patients (nearly 85%) suffer TSS repeatedy without ever mounting a humoral immune response to TSST-1 (Schlievert 1993), and TSS patients exhibit reduced IgG serum levels (Bohach et al. 1990).  Superantigenic T cell activation by TSST-1 appears to stimulate widespread production of interferon (IFN) g by T lymphocytes.  IFN-g has been shown to inhibit Ig production and could explain the observed humoral immunosuppression: B cells are not stimulated to produce Ig and never mount a primary immune response (Schlievert 1993).  This theory seems to be widely accepted, though in mice IFN-g actually stimulates isotype switching to IgG3 and IgG2a (Janeway et al. 1999).

Lack of Purulence.  The widespread release of TNF-a seems to inhibit the chemotactic capabilities of neutrophils and no localized inflammation or purulence (pus formation) is seen in the area of initial bacterial infection (Bohach et al. 1990).

Necrotizing Fasciitis (TSLS only).  Streptococcus pyogenes infection usually causes severe, gangrenous necrosis of subcutaneous fascia and fat tissue, though the surrounding skin and muscle is usually unharmed (Stevens 1995).

Figure 2.  Summary of causes and effects of TSS and TSLS.  Solid lines indicate known effects, broken lines indicate putative relationships.  Image created by author.

Because no lab tests are available to confirm diagnosis of TSS or TSLS, clinicians must rely on the following standard case definitions (Chesney 1989).

Staphylococcal TSS (Deresiewicz 1999):

Streptococcal TSLS (Stevens 1995):

A. Isolation of group A Streptococcus
    1. from a sterile site
    2. from a nonsterile site
B.  Clinical signs of severity
    1. Hypotension
    2. Clinical and laboratory abnormality (2 or more of following):
        a. renal impairment
        b. coagulopathy
        c. liver abnormalities
        d. acute respiratory distress syndrome
        e. necrotizing fasciitis
      f. erythrematous rash

Definite case = A1 + B(1 + 2)
Probable case = A2 + B(1 + 2)

Treatment for both TSS and TSLS involve the same basic steps:

1) Clean any obvious wounds and remove any foreign bodies (Chesney 1989).

2) Prescription of appropriate antibiotics to eliminate bacteria (Chesney 1989); Stevens (1985) recommends penicillin and clindomycin for S. pyogenes.

3) Monitor and manage all other symptoms, e.g. administer IV fluids (Chesney 1989).

4) For severe cases, administer methylprednisone, a corticosteriod inhibitor of TNF-a synthesis (Bohach et al. 1990).

5) For TSLS, debride necrotizing fascia in timely manner; perform fasciotomy or amputation as necessary (Stevens 1995).

One possible treatment for severe TSS is IV delivery of pooled anti-TSST-1 IgG.  IgG treatment will neutralize TSST-1 toxin, but it will also inhibit the production of a native primary immune response and the formation of immunological memory for TSST-1 (Chesney 1989) due to the antibody-mediated suppression of naive lymphocyte activation (Janeway et al. 1999).  Therefore, IV Ig treatment is only recommended for severe or recurrent cases of TSS (Chesney 1989).

As of 1989, no antitoxins were recommended for use in TSS patients (Chesney 1989), but in April 2000 several Isreali researchers reported the development of an agonist peptide for TSST-1 that prevented TSS in 100% of mice when administered before TSST-1 inoculation and rescued 50% of mice from TSS when administered after onset of TSS (Siegel-Itzkovich 2000).  Of course, clinical trials will be needed to determine the effectiveness of such agonists in treating human TSS.

Bohach GA, Fast DJ, Nelson RD, Schlievert PM.  1990.  Staphylococcal and streptococcal pyrogenic toxins involved in toxic shock syndrome and related illnesses.  Critical Reviews in Microbiology 17(4): 251-272.

Chesney PJ.  1989.  Clinical aspects and spectrum of illness of toxic shock syndrome: Overview.  Reviews of Infectious Diseases 11(supplement 1): S1-S7.

Deresiewicz RL.  1999.  Toxic shock syndrome: a health professional's guide.  <http://www.toxicshock.com/healthprof.htm>  Accessed 2000 20 April.

Forni AL, Kaplan EL, Schlievert PM, Roberts RB.  1995.  Clinical and microbiological characteristics of severe group A streptococcus infections and streptococcal toxic shock syndrome.  Clinical Infectious Diseases 21: 333-340.

Janeway CA, Travers P, Walport M, Capra JD.  1999.  Immunobiology: the immune system in health and disease.  4th ed.  New York: Elsevier Science, Ltd/Garland Publishing.

Kum WWS, Laupland KB, Chow AW.  2000.  Defining a novel domain of staphylococcal toxic shock syndrome toxin-1 critical for major histocompatibility complex class II binding, superantigenic activity, and lethality.  Canadian Journal of Microbiology 46(2): 171-179.

Schlievert PM.  1993.  Role of superantigens in human disease.  Journal of Infectious Diseases 167: 997-1002.

Schmitt CK, Meysick KC, O'Brien AD.  1999.  Bacterial toxins: friends or foes?  Emerging Infectious Diseases 5(2).  <http://www.cdc.gov/ncidod/_vti_bin/shtml.dll/EID/vol5no2/schmitt.htm/map>  Accessed 2000 20 April.

Siegel-Itzkovich J.  2000.  Scientists build a peptide to stop toxic shock syndrome.  British Medical Journal 320: 958.  <http://www.bmj.com/cgi/content/full/320/7240/958>  Accessed 2000 20 April.

Stevens DL.  1995.  Streptococcal toxic-shock syndrome: Spectrum of disease, pathogenesis, and new concepts in treatment.  Emerging Infectious Diseases 1(3) 69-78.