The occurrence of TSS is linked to the bacterial secretion of superantigenic exotoxins, which are bacterial virulence factors genetically encoded and secreted. Superantigenic exotoxins are able to induce unconventional activation of T cells by antigen-presenting cells (APCs).

During conventional T-cell activation, the APC absorbs foreign particles, processes protease digestion, and presents them as partially degraded in a specific binding groove in the major histocompatibility complex class II (MHC II), which is expressed on its surface. The Ag-MHC II complex binds to the surface of the T-cell receptor (TCR). This results in monoclonal activation of T cells specific to the antigen (Ag).

In the TSS, the superantigen binds the TCR and MHC II outside the Ag presentation site with high affinity. This results in nonspecific, polyclonal lymphocyte activation of 5 to 30% of the total population of T cells [11,12,13]. This simultaneous polyclonal activation results in a significant activation of NF kappa B, which plays a major role in the generation and expansion of the inflammatory response [1]. This results in a massive release of proinflammatory cytokines, with clinical signs, such as capillary leakage, arterial hypotension, organ failure, and coagulation activation, usually being reported in this setting [1].

Physiopathological specificities of staphylococcal and streptococcal TSS are detailed in the corresponding subparts.

The first description of this syndrome was published by James Todd and colleagues in The Lancet in 1978 [14]. The authors described a pediatric case series of seven children, with clinical presentations including high fever, cephalalgia, confusion, cutaneous rash, conjunctival hyperhemia, and digestive signs. The children progressed to a state of prolonged severe shock associated with renal and hepatic failure and disseminated intravascular coagulation. Exotoxin-producing S. aureus was isolated from the foci of infection (empyema and abscess) in two patients and in mucosal swabs (nasopharyngeal, vaginal, and tracheal) in four patients but not from blood, cerebrospinal fluid (CSF), or urine. One patient died, while all the others survived and presented with desquamation of the palm of the hands or sole of the feet during recovery [14]. Staphylococcal TSS in adult patients was then described in the 1980s and predominantly involved menstruating women [15].

The diagnostic criteria for staphylococcal TSS were proposed by the Centers for Disease Control and Prevention (CDC) in the 1980s and revised in 2011 [1,16]. These criteria, combining clinical and laboratory aspects, are presented in Table 1.

However, CDC criteria only allow for a retrospective diagnosis, as they include the desquamation of the palms of the hands and soles of the feet, which occurs 8 to 21 days after the onset of the illness [1,12,17]. Moreover, a French multicentric retrospective study describing 102 cases of staphylococcal menstrual TSS (m-TSS) demonstrated that none of them met the CDC criteria for a confirmed TSS, and only half of them met the criteria for a probable TSS [18].

Staphylococcal TSS is rare. According to recent studies, the annual incidence of TSS is estimated to be between 0.03 and 0.07/100,000 population [19,20] and seems to be stable. A peak in incidence (13.7/100,000 persons) was observed in the 1980s in the USA [21], linked to the use of highly absorbent tampons, but its incidence decreased after changes in tampon manufacture [22]. These features explain the differentiation in the literature of staphylococcal TSS between menstrual (m-TSS) and nonmenstrual (nm-TSS) syndromes. While m-TSS cases were largely predominant in the 1980s, compared to nm-TSS cases, the proportion of nm-TSS cases gradually increased over time [19,22]. In the UK, the incidence of m-TSS was estimated to be 0.09/100,000 and that of nm-TSS was estimated to be 0.04/100,000 persons [19]. The highest incidence of m-TSS (1.41/100,000 persons) is observed in women aged 13 to 24 years [23].

Menstrual toxic shock syndrome (m-TSS), which usually occurs in healthy young menstruating women [23], is linked to vaginal colonization with toxic shock syndrome toxin—1 (TSST-1)-producing S. aureus in women without neutralizing antibodies. An American study on a cohort of 262 women showed that between 2003 and 2005, 22.9% of women were vaginally colonized with S. aureus, and 4.2% were colonized with TSST-1-producing S. aureus during both menstruation and nonmenstruation [24]. The use of tampons creates a physicochemical environment favorable to S. aureus’ growth and production of TSST-1, in particular by providing oxygen in this anaerobic medium [25]. TSST-1 can bind to vaginal epithelial cells and cross the vaginal mucosa [13,25]. A large majority of m-TSS patients have undetectable levels of protective antibodies at the onset of the illness [26]. In m-TSS, blood culture results are negative in all the publications [16]. As it occurs in the absence of any staphylococcal infection, m-TSS is an exclusively toxin-mediated shock.

In a French multicentric retrospective (2005–2020) study reporting 102 cases of m-TSS, the median age was 18 (16–24) years. No previous comorbid condition was reported in 87% of the cases. Clinical presentation included tachycardia (median heart rate 128 (115–140)/min), high fever (median temperature 39.4 (38.5–40.0) °C), skin rash (87% of the cases), and mucosal involvement (50% of the cases). Digestive signs (abdominal pain, diarrhea, and vomiting) and cephalalgia were very common [18]. Vasopressor support was needed in 84% of the cases, 21% of which needed mechanical ventilation [18]. In this study, all patients were using tampons during their period [18]. However, vaginal cups and intrauterine devices have also been reported in staphylococcal m-TSS [27,28,29,30].

Nm-TSS can result from any staphylococcal infection with a toxin-producing strain of S. aureus. It is most often postoperative, even after relatively simple procedures, but can occur postpartum, after abortion, or because of nonsurgical cutaneous lesions [22]. All types of surgical procedures can precede postoperative TSS, but plastic, orthopedic, and head and neck surgery are most frequently used [12]. Postoperative TSS occurs after a median delay of 4 days after surgery [12]. Blood cultures are positive for S. aureus in 50% of cases [31]. As nm-TSS is linked to a staphylococcal infection, it is a mixed septic and toxin-mediated shock.

The clinical presentation of nm-TSS is highly comparable to that of m-TSS, although it occurs in significantly older patients (33 (0–84) years vs. 19 (10–47) years, p = 0.008) [31]. A retrospective (2003–2006) multicentric French study compared the clinical presentation of m-TSS (21 cases) and nm-TSS (34 cases). Digestive signs and mucosal involvement were less frequent (74% vs. 100%, p = 0.009, and 42% vs. 76%, p = 0.024, respectively), but neurological involvement was significantly more frequently observed (61 vs. 29%, p = 0.028). No statistically significant difference was observed between m-TSS and nm-TSS in the occurrence of other CDC clinical criteria (fever, rash, desquamation, hypotension, renal, hepatic, and hematologic failure) [31]. Another retrospective (2000–2006) American study compared m-TSS (33 cases) and nm-TSS (28 cases) and reported no difference in clinical presentation [23].

The described staphylococcal superantigenic exotoxins include TSST-1 and enterotoxins (of which approximately thirty have been described to date) [32]. TSST-1 is a 194 amino acid protein encoded by the gene tst and is responsible for 89 to 95% of m-TSSs and 50% of nm-TSSs, the other half being related to the secretion of staphylococcal enterotoxins A, B, and C [23,25,33,34]. M-TSS are nearly all linked to the USA200 clonal group [25].

S. aureus strains are, in a vast majority of cases, methicillin-susceptible strains. In a recent French multicenter retrospective study (2005–2020) describing 102 cases of m-TSS, no case of methicillin-resistant S. aureus (MRSA) was described [18]. The UK national surveillance data on 180 TSS (107 nm-TSS) between 2008 and 2012 identified only 7 (3.8%) nm-TSS cases of MRSA isolates [19]. However, 4 cases (7%) of MRSA isolates were identified in an American study (2000–2006) on 61 staphylococcal TSS—2 of them being suggestive of community-associated MRSA and 1 of them being USA400 MRSA [23]. Case reports of nonmenstrual staphylococcal TSS involving MRSA have also been published [35,36,37].

Nasal colonization with TSST-1-producing S. aureus could be a risk factor for postoperative TSS. S. aureus nasal colonization has been observed in 20 to 80% of the human population [38,39] and identified as a major risk factor for community-acquired and nosocomial infections. A recent prospective multicenter study confirmed that preoperative S. aureus carriage in nose, throat or perineum, was associated with both surgical site infections and bloodstream infections [40]. Unfortunately, this study did not provide any information on TSST-1 production by the S. aureus isolates. A recent study analyzing nasal colonization of 150 healthy volunteers in Kabul showed that 68.4% of the MRSA isolates were TSST-1 producers. TSST-1 production by MSSA isolates was not reported in this study [41]. Nasal colonization with TSST-1-producing S. aureus in patients with staphylococcal TSS have not been specifically assessed. Most of the published studies did not report nasal carriage of S. aureus. Celie et al. have reported positive S. aureus nasal cultures in some cases of postoperative TSS, but these samples were collected in the operating site in all cases [12]. To date, the incidence of nasal S. aureus colonization in staphylococcal TSS is unknown.

In 1987, Cone et al. described, in the New England Journal of Medicine, cases of two patients with severe GAS infection with a clinical presentation similar to staphylococcal TSS. This syndrome was named “streptococcal toxic shock-like syndrome” [42]. Two years later, Stevens et al. reported a case series of 20 patients with severe GAS infection, with clinical presentations including shock, multiorgan system involvement, and rapidly progressive local tissue destruction [43].

The diagnostic criteria from the CDC for streptococcal TSS [1,44] are presented in Table 1. They include clinical signs of severity associated with the presence of GAS in a nonsterile site (throat, vagina, and sputum) or a normally sterile site (CSF, blood, peritoneal fluid, and tissue biopsy) [1].

Streptococcal TSS is rare. Overall, 8 to 22% of patients with severe S. pyogenes infection will develop streptococcal TSS [45,46,47,48]. Approximately 40 to 50% of patients with necrotizing soft tissue infection (NSTI) will develop streptococcal TSS [46,47]. Blood cultures are positive in 60 to 86% of cases [6,49]. Common sources of infection include the vagina, pharyngeal mucosa, skin and soft tissues. Streptococcal TSS can also complicate a minor trauma without skin effraction, pneumonia, intrauterine device, septic arthritis, burn, chickenpox in children, or occur postpartum in young women [46,50]. The source of infection remains unknown in 50% of cases [49].

Streptococcal TSS mostly occurs in elderly patients between 50 and 69 years of age and in patients with comorbidities (diabetes, malignancy, hepatic disease, chronic renal impairment, and heart disease) [45,48]. Alcoholism and the use of nonsteroidal anti-inflammatory agents (NSAIDs) have also been suspected to be risk factors for streptococcal TSS [46]. A strong association between the use of NSAIDs and occurrence of necrotizing soft tissue infection has been described [51]: there is a 3-fold increased risk for streptococcal TSS [47]. However, the role of NSAIDs in streptococcal TSS remains debated. An experimental study in a murine model showed that NSAIDs’ administration resulted in a 22-fold increase in the number of GAS in an injured muscle [52]. Administration of NSAIDs could also mask the signs of severity of the infection by attenuating inflammatory signs, and delaying the diagnosis with a negative impact of the prognosis.

Clinical presentation of streptococcal TSS was described in a series of 14 cases in North Yorkshire. Hypotension was described in 100% of cases, acute kidney failure in 93%, liver failure in 57%, and disseminated intravascular coagulation in 64% [6]. Multiorgan failure syndrome was reported in 43% of the cases [6]. In this series, streptococcal TSS was associated with a necrotizing infection in 71% of the cases (predominantly NSTI and myonecrosis) [6].

The disease occurs after penetration of the exotoxin-producing S. pyogenes through a skin or mucous barrier alteration. S.pyogenes then spreads to deep tissues. The main superantigenic exotoxins described in S. pyogenes are streptococcal pyrogenic exotoxins (SpE) A, B, and C and streptococcal superantigen A (SsA). The majority of streptococcal isolates causing TSS are the emm1 (41.1% of the cases), emm3 (8.4% of the cases), emm28 (8.9% of the cases), and emm89 (9.8% of the cases) genotypes [45]. Streptococcal TSS occurs more frequently with GAS strains harboring SpeA or Spec genes (p ≤ 0.001) than those harboring Ssa genes [45]. SpeB participates in the rapid dissemination of S. pyogenes in the skin and soft tissues in combination with other streptococcal virulence factors, such as soluble M protein, which participate in the local and systemic excessive activation of T lymphocytes, APCs and neutrophils.

Clinical case reports of TSS involving various bacteria (group B, C and G streptococci, Yersinia pseudotuberculosis, Pseudomonas fluorescens, Mycoplasma arthritidis, Clostridium, and coagulase-negative staphylococci (CNS)) have been reported [2,3,4,5,6,7,8,9,10]. The pathophysiology of these probable TSS is not yet well established. Group B and G streptococci produce pyrogenic toxins that are able to induce lethal endotoxin shock in animals [4,5]. Superantigen production has also been documented for Mycoplasma arthritidis (Mycoplasma arthritidis-derived superantigen) [53] and Yersinia pseudotuberculosis [54]. Previous studies have shown contradictory results regarding the ability of the CNS to produce superantigens [55,56,57,58]. However, stimulation of human monocytes by the killed CNS could induce a dose-dependent production of cytokines responsible for the clinical symptoms [59]. To date, only a few cases of these TSS have been reported in the literature, and additional data are therefore needed.