INVASIVE GROUP A streptococcal (GAS) infections include bacteremia, deep soft tissue infection (necrotizing fasciitis and myonecrosis), pneumonia, empyema, post-partum sepsis, severe scarlet fever (see below), meningitis, peritonitis, and septic arthritis. Most of these infections have been described in the Latin, German, and English literature for several centuries and should not be considered new entities. By the late 1800s and early in the 1900s, but before the availability of antibiotics, most of these types of infection became milder illnesses. However, by the early to mid 1980s, individual case summaries and then larger reports were published that described fulminant group A streptococcal infections now known as streptococcal toxic shock syndrome (StrepTSS). StrepTSS is essentially defined as the rapid onset of shock and organ failure associated with any acute group A streptococcal infection. Thus, a patient may have StrepTSS secondary to primary bacteremia, necrotizing fasciitis, myonecrosis, post-partum sepsis, or pneumonia, to mention a few. During the course of the last 15 years clinicians around the world have gained valuable experience in the diagnosis and management of such severe invasive group A streptococcal infections. Despite such knowledge there are many aspects of these diseases that remain unknown and even mysterious. This mini-review will briefly describe what is and is not known, and what is controversial.
Streptococcal Toxic Shock Syndrome: Epidemiology of Primary and Secondary Cases
Population-based studies have documented an annual incidence of 3.5 cases of invasive GAS infection per 100,000 population, with most cases being sporadic in nature. However, higher rates have been described in discrete areas. In 1994, an epidemic of related invasive infections occurred in Winnamango, Minnesota with an annualized prevalence of 24 cases per 100,000 population . In Missoula, Montana in 1999, the incidence of invasive infections reached 30 cases per 100,000 population (author’s unpublished observations).
Studies conducted over nearly 100 years have clearly established that Group A streptococcus is quickly and efficiently transmitted from an index case to susceptible individuals. Transmission may result in asymptomatic colonization or actual infection such as pharyngitis, scarlet fever, rheumatic fever, or post streptococcal glomerulonephritis. Invasive Group A streptococcal infections have also been described in hospitals, convalescent centers [2,3] and among hospital employees and family contacts of patients with invasive infections. Some of these studies have documented identical strains (i.e., the same M-type and identical RFLP pattern) from primary and index cases. In addition, carriage of Group A streptococcus by healthcare personnel has been associated with the spread of life threatening Group A streptococcal infections in the obstetrics/gynecology and ear/nose/ throat wards of American hospitals . Such infections have also originated in outpatient surgical settings and within the home environment.
The risk of developing an invasive infection after contact with a primary case may be as high as 50–60 times greater than the general population. Active surveillance by the Centers for Disease Control and Prevention has documented 1 invasive case among more than 1,500 contacts of an invasive case . Transmission is clearly related to duration of close or intimate contact and crowding, as well as host factors such as 1) active viral infections such as varicella or influenza; 2) recent surgical wounds and childbirth (author’s unpublished observations); 3) absence of type specific opsonic antibody against the Group A streptococcus causing the index case; and 4) absence of neutralizing antibody against pyrogenic exotoxin A or B.
Acquisition of Group A Streptococcus
The known portals of entry in invasive streptococcal infections are the vagina, pharynx, and skin (Table 1). Interestingly, a defined portal cannot be established in 50% of cases . Only rarely do patients with symptomatic pharyngitis develop StrepTSS. Occasionally, pneumonia due to GAS results in bacteremia and StrepTSS. Surgical procedures such as suction lipectomy, hysterectomy, vaginal delivery, bunionectomy, and bone pinning have provided a portal of entry in some cases. In patients with StrepTSS without a portal of entry, infections have developed within 24–72 hours at the site of antecedent minor non-penetrating trauma resulting in hematoma or muscle strain . In these patients, streptococci likely seed traumatized deep tissues via a transient bacteremia, and once within this environment, bacterial growth begins, causing systemic signs and symptoms of disease. In this setting, cutaneous findings are late manifestations of infection. Early diagnosis in these cases is much more difficult than in those patients with a defined portal of entry; however, severe localized pain and fever are the best clues. Virus infections such as varicella and influenza likely provide portals of entry in other cases . In some cases the use of nonsteroidal anti-inflammatory agents may mask the presenting symptoms, attenuate signs and symptoms of infection, and may also predispose individuals to more severe streptococcal infection and shock .
Twenty-percent of patients have an influenza-like syndrome characterized by fever, chills, myalgia, nausea, vomiting, and diarrhea (Table 2) . Fever is the most common presenting sign, although hypothermia may be present in patients with shock . Confusion is present in 55% of patients, and in some, coma or combativeness are present . Pain–the most common initial symptom of StrepTSS–is abrupt in onset, severe , and usually precedes tenderness or physical findings. The pain usually involves an extremity but may also mimic peritonitis, pelvic inflammatory disease, pneumonia, acute myocardial infarction, or pericarditis .
Eighty-percent of patients have soft tissue infection and about 50% of these progress to necrotizing fasciitis or myositis . Of the 20% of cases without soft tissue findings, a variety of clinical presentations were observed, including endophalmitis, perihepatitis, peritonitis, myocarditis, septic joint, post-partum sepsis, empyema, meningitis, and overwhelming sepsis . A diffuse, scarlatina-like erythema is uncommon, occurring in only 10% of cases.
Laboratory Evaluation of Patients with StrepTSS
The serum creatinine phosphokinase (CPK) level is useful in detecting the presence of deeper soft-tissue infections, and when the level is elevated or rising, there is a good correlation with necrotizing fasciitis or myositis . The white blood cell count may be low, normal, or elevated; however, the mean percentage of immature neutrophils (including band forms, metamyelocytes, and myelocytes) is striking, reaching 40–50% . Therefore, in many settings, a manual differential count is very important. Hemoglobinuria and elevated serum creatinine values are evidence of renal involvement. It is important to note that renal impairment precedes hypotension in 40–50% of cases . Hypoalbuminemia and hypocalcemia occur early and become profound within 24–48 hours of admission. Patients with StrepTSS, particularly those with bacteremia, may have brisk hemolysis presumably due to the hemolysins, streptolysin O, and streptolysin S.
The rapidity with which shock and multi-organ failure can progress is impressive and many patients may die within 24–48 hours of hospitalization . Shock was apparent at the time of admission or within 4–8 hours in virtually all patients. In only 10% of patients did systolic blood pressure became normal 4–8 hours after administration of antibiotics, albumin, and electrolyte solutions containing salts or dopamine; in all other patients shock persisted. Interestingly, renal dysfunction (mean serum creatinine > 2.4 mg/dL) preceded shock in most cases and was apparent on admission. Renal failure progressed or persisted in all patients for 48–72 hours, and several patients required dialysis for 10–20 days . In patients who survived, serum creatinine values returned to normal within 4–6 weeks. Acute respiratory distress syndrome (ARDS) occurred in 55% of patients and became apparent after the onset of hypotension . The severity of ARDS was such that supplemental oxygen, intubation, and mechanical ventilation were necessary in over 90% of patients who developed this syndrome . Mortality rates have varied from 30–70% [6,8,9]. However, morbidity is also high, and over 60% of patients in one series underwent major surgical procedures, which included fasciotomy, surgical debridement, exploratory laparotomy, intra-ocular aspiration, amputation, or hysterectomy .
Poor prognostic indicators include: 1) persistent hypotension despite IV fluids (colloid, crystalloid, blood, and albumin); 2) worsening metabolic acidosis; 3) disseminated intravascular coagulopathy; and 4) necrotizing infections not amenable to surgical debridement (e.g., infections of the head, neck, mediastinum, pelvis, or infections in patients considered high risk for surgery;Table 3). Experimentally in nonhuman primates, a dropping white count in the face of septic shock was associated with increased mortality and more severe necrotizing fasciitis .
Is StrepTSS a New Disease or are Modern Strains of GAS More Virulent?
There is little question that in previous centuries patients probably developed shock and organ failure in association with a variety of streptococcal infections. For example, soft tissue infections associated with bacteremia have been described earlier in the last century in association with nursing home patients, and this clinical entity persists into the current century (reviewed in ). In addition, post-partum sepsis was common in the mid 1850s in Vienna as well as the United States, and these infections had a high mortality. Interestingly, such infections disappeared throughout the world for over 100 years, merely due to better birthing practices. This dramatic reduction documents the effectiveness of infection control practices, such as hand washing, gloves and simply recognizing the potential of spreading contagion within the hospital by health care providers. The re-emergence of post-partum sepsis during the last 15 years in the face of such practices is remarkable and deserves an explanation. Most likely, this re-emergence is due to enhanced virulence of the strains of streptococcus existing in communities worldwide for the last 15 years. While it is true that recent sporadic cases and even some epidemics have been traced to hospital personnel who are rectal, vaginal, or pharyngeal carriers of GAS, it seems obvious that contamination of the birthing field with even a small number of these modern GAS strains is sufficient to cause infection leading to shock, organ failure, uterine necrosis, and even death. In stark contrast, during the childbed fever era of the 1850s, millions of viable GAS infections were likely transferred from the infected uteruses of women who died of post-partum sepsis to the birth canals of pregnant women by the unwashed hands of obstetricians. Thus, it seems entirely logical, albeit hypothetical, that fewer organisms are sufficient to cause post-partum sepsis in the modern era compared with the days of Semmelweis and Oliver Wendell Holmes. By inference, these contemporary strains of GAS must be inherently more virulent.
Is Necrotizing Fasciitis Today the Same as that Described in 1924 by Meleney?
Similarly, necrotizing fasciitis caused by hemolytic streptococci was first described nearly 100 years ago by Meleney while in China,  and he called this entity streptococcal gangrene. His clinical descriptions of the temporal progression of necrotizing fasciitis remain classical. However, there are major differences between what he described and the necrotizing fasciitis of today. First, he described the progression of necrotizing fasciitis over the course of 10–14 days, in which redness of the skin gives way to bluish discoloration, bullae formation, and later blackish discoloration with sloughing of skin. While most infectious disease physicians during the last 15 years would agree with Meleney’s clinical descriptions, the course of modern streptococcal gangrene is compacted in time and generally runs its course in only 48–96 hours. This temporal discrepancy becomes more striking when one considers that patients of Meleney’s era received only fasciotomy and the rinsing of the open wounds with Dakan’s solution; his descriptions were written well before the advent of antibiotics. Today, aggressive surgical debridement is also necessary to ensure survival for patients with invasive streptococcal soft tissue infections. However, despite our modern advantages of antibiotics, better surgical procedures, ventilator support, renal dialysis, intravenous fluids, etc., the mortality rate for such infections is higher now (30–70%) than in Meleney’s era (20%). Thus, the more fulminant nature of necrotizing fasciitis (StrepTSS) today is most likely caused by strains of GAS having enhanced virulence compared to strains causing infections during the mid 19th century.
Is StrepTSS Merely a Severe Form of Scarlet Fever?
There are clearly septic and toxic forms of scarlet fever (reviewed in ), yet these entities differ from StrepTSS in several ways. First, every description of septic and toxic scarlet fever in the literature over the course of the last 150 years has been in children. In contrast, StrepTSS occurs in all age groups, and is in fact more common in adults. Scarlet fever in general is very rare in adults and has its highest prevalence in children from 4–10 years of age. Second, septic scarlet fever was described in children who first had pharyngitis and the typical scarlatina rash, and then developed progression of infection to the sinuses, mastoids, middle ear, meninges, or vital structures of the neck. Such patients frequently died agonizing deaths 2–4 weeks into their illness after the local extension of infection resulted in airway obstruction, erosion into a major vessel of the neck, etc. Apparently, these patients did not develop shock and organ failure early in the course of illness as do patients with StrepTSS. Patients with toxic scarlet fever did develop severe hyperpyrexia, intractable seizures, and sometimes died within several days. However, if antipyretics, antibiotics or anti-seizure medications had been available, these deaths may have been preventable. Some patients with typical scarlet fever also died late in the course of the illness of post infectious sequelae such as rheumatic fever and post-streptococcal glomerulonephritis; however, these complications developed well after the symptoms and signs of scarlet fever had resolved. Thus, these severe forms of scarlet fever differed from StrepTSS in that in bacteremia was uncommon, age predilection was children, shock and organ failure were uncommon, and necrotizing fasciitis was rare (Table 4).
Are the Same Toxins Responsible for both Scarlet Fever and StrepTSS?
There is considerable evidence that the pathogenesis of both scarlet fever and StrepTSS involves the streptococcal pyrogenic exotoxins, also called erythrogenic and scarlatina toxins. In the 1920s, George and Gladys Dick clearly demonstrated that scarlet fever was caused by extracellular protein toxins [13,14]. It was also shown that antibodies (or at least unpurified sera) from children recovering from scarlet fever were protective and attenuated the course of scarlet fever in children [15,16]. The observations that some children had several episodes of scarlet fever led to the discovery of pyrogenic exotoxins A, B, and C. In the last decade several other pyrogenic toxins have been described as well. Numerous studies have implicated pyrogenic exotoxins in the pathogenesis of StrepTSS (reviewed in ), yet no single pyrogenic exotoxin has been demonstrated in all cases. Thus, a variety of pyrogenic toxins individually or in combination may contribute to modern invasive disease. In fact, seven different pyrogenic exotoxins and multiple M-types of Group A streptococci have been isolated from patients with invasive streptococcal infections such as StrepTSS, necrotizing fasciitis, and bacteremia. Alternatively, other Group A streptococcal virulence factors may be involved in the pathogenesis of StrepTSS. For instance, we have demonstrated that all recent isolates of M-1, the most common strain of GAS associated with StrepTSS, produce a toxin called NADase . Interestingly M-1 strains isolated from 1970–1985 (i.e., before the recent emergence of StrepTSS) did not produce this toxin. Additionally, 100% of all other M-types associated with StrepTSS (M-3, M-11,M-12, M-4, M-6 and M-28) also produce NADase. Thus, NADase production has the highest correlation of any single GAS virulence factor with invasive infections. The role that NADase plays in pathogenesis of invasive infections or scarlet fever has not been established, though in vitro this enzyme has profound effects on neutrophil function .
The role of the pyrogenic exotoxins in pathogenesis is likely related to their ability to act as superantigens, simultaneously activating both antigen presenting cells (e.g., monocytes, macrophages) and lymphocytes. This results in high level production of proinflammatory cytokines (TNFα, IL-1β, and IL-6) as well as lymphokines (interferon-γ and TNFβ). Other GAS cell-associated components (e.g., peptidoglycan, LTA, M-protein) and extracellular toxins (streptolysin O) may contribute to the generation of cytokine production in vivo. This is particularly true in those patients with StrepTSS who are bacteremic (60%).
That cytokines play an important role in StrepTSS is demonstrated in recent experimental studies in non-human primates in which infusion of a monoclonal antibody against TNFα improved outcome and reduced organ dysfunction in animals with group A streptococcal bacteremia .
Strategies to Prevent or Modify the Course of StrepTSS.
The appropriate treatment of infection caused by highly virulent GAS strains is quite dependent upon the stage of clinical illness, but because of the fulminant nature of these infections, the clinician has a very short time to initiate therapy. Thus, a high index of suspicion, timely ordering and interpretation of diagnostic tests, and early consultations with infectious disease specialists and surgeons is crucial. In cases where StrepTSS occurs following surgical procedures, it is clear that some patients may acquire infection, develop shock and organ failure, and die - all within 72 hours. The following stages of GAS infection provide some clues as to therapy, pathogenesis and clinical signs.
Stage I (Table 1) represents acquisition of GAS from 1) the external environment, such as contact with a primary case of some type of GAS infection (pharyngitis, StrepTSS etc.); 2) contact with a carrier of GAS; or 3) from endogenous sources (throat or skin). It is likely that infection will not develop unless there are predisposing factors, such as a portal of entry (burns, surgical incision, insect bite, varicella, recent child birth) or blunt trauma (hematoma, muscle strain, joint effusion) that provide a niche deep in tissue for hematogenously delivered GAS to proliferate. At this stage a wide variety of antibiotics including penicillin, cephalosporins, macrolides, or clindamycin would be effective in treating, arresting, or preventing infection largely because this is an early stage and the number of streptococci is small. Prophylactic treatment of contacts of a known case should be considered, especially if the contact has predisposing factors such as those mentioned above. If several cases of GAS have occurred in a health care setting, epidemiologic studies should be performed to determine if a health care provider is the source of the outbreak. Since numerous reports have documented such epidemics over the course of several months, even years, the source is usually a health care worker who is an asymptomatic carrier of GAS. Carriage may occur on skin, pharynx, vagina or rectum. Experience in treating pharyngeal carriers in households of patients with rheumatic heart disease suggests that the “carrier state” is best treated with penicillin plus rifampin.
Stage II is also an early stage of infection where the number of organisms is probably in the 102–103 range and toxin production has begun, but in small quantities. Patients may experience modest local signs of inflammation but have few systemic signs of infection (Table 1). Diagnosis of infection is difficult at this stage, unless such patients have a visible portal of entry, in which case modest redness, swelling, and tenderness may be apparent. Patients with no portal who have deeper infection from hematogenous seeding of traumatized tissue may have only pain that progressively becomes more severe. These patients may respond well to beta-lactam antibiotics presumably because organisms are in the logarithmic phase of growth, but macrolides or clindamycin would likely be superior due to their ability to suppress both bacterial growth and toxin production. No direct comparison studies have been performed in patients in this stage of infection.
Stage III is later in the course of infection where organisms have reached the stationary phase of growth and toxin production is maximal. Patients with a portal of entry will have marked redness, swelling, tenderness, and heat. There may be drainage from the initial portal of entry. Patients without a portal who have deep infection will have excruciating pain and tenderness, with or without redness or swelling. Both groups of patients will have systemic evidence of infection such as fever, chills, nausea, vomiting, and both will have tachycardia, laboratory evidence of marked left shift, increased creatinine, hypotension, CPK elevation, low albumin, and low calcium. Over half of such patients will have bacteremia. Beta lactam antibiotics are not efficacious at this stage because organisms are in stationary phase and penicillin binding proteins are not fully expressed . In contrast, clindamycin is unaffected by the physiologic state of the organism and would effectively inhibit bacterial growth and suppress toxin production. In addition to these activities, clindamycin is also a potent suppressor of TNF production by human monocytes at readily achievable concentrations of drug (5 μg/ml) [20,21]. For these reasons, it is not surprising that clindamycin has been shown to be superior to beta-lactam antibiotics in experimental  and human cases of severe invasive streptococcal infection .
Intensive care monitoring and support are crucial at Stage III and intravenous fluids are required in massive quantities. Among toxic patients with either superficial or deep soft tissue infection, radiographic studies and surgical intervention may be necessary to inspect the fascia and muscle, and, if necrosis is found, to debride appropriately. Surgical debridement probably also removes large concentrations of toxins and cytokines. In addition to toxin suppression, neutralization of circulating toxins with intravenous gamma globulin (IGIV) [24,25] may also be associated with improved survival . In the latter study , IVIG therapy reduced the mortality in 15 patients compared with 15 historical controls. Patients in the IVIG arm were more likely to have received clindamycin and were more likely to have surgical debridement performed than the historical controls. Note: A double blind study investigating the efficacy of IGIV in patients with StrepTSS is underway in Sweden. While the efficacy of anti-cytokine strategies has not been studied in humans, recently a neutralizing monoclonal antibody against TNF was efficacious in a primate model of GAS bacteremia . Other potential strategies to reduce toxins in the circulation include hemodialysis, hemoperfusion and plasmaphoresis. Dialysis has been performed in large numbers of patients with StrepTSS and plays an important role in management, though there is no data to demonstrate removal of toxins by this method. In general dialysis and hemoperfusion effectively remove molecules of molecular mass <500 daltons, however most bacterial protein toxins have masses of 18–50,000 daltons. Hemofiltration shows more promise as more porous membranes are used and molecules of the size of bacterial toxins and cytokines can be removed. Plasma exchange has been used in StrepTSS and appeared to reduce the duration of time patients needed hemodialysis, though the authors did not state whether mortality and morbidity had been reduced by this method .
Stage IV is very late in the disease course and is characterized by high levels of circulating toxins and cytokines. If necrotizing fasciitis or myonecrosis is present, overlying soft tissues will demonstrate marked swelling, a purplish color of the skin with blisters or bullae, and sloughing of the skin. Profound hypotension, multiple organ dysfunction, disseminated intravascular coagulopathy, and ARDS are usually present. These patients have a high mortality despite appropriate treatment as defined above. Improvement of mortality at this stage may require a combination of strategies to neutralize both toxins and cytokines, aggressively debride tissue and specifically address the disseminated coagulopathy.
Invasive GAS infections have been described for many centuries, yet in the mid-1980s the severity of such infections increased dramatically. Careful epidemiologic studies and case reports have documented that these types of infections affect all age groups in all regions of the world. In the absence of both racial predispositions and a predilection for immunocompromised hosts, we attribute the occurrence of these severe GAS infections to increased virulence of the organism itself. The relatively low incidence of these infections, 3.5 cases/100,000 population per year, despite a high prevalence of “virulent strains” (M-1, M-3, M-6, M-28, etc.) in the throats of children and adults in all communities suggests that increased virulence alone is not sufficient to explain this phenomenon. Clearly, other factors are required in most individuals to develop severe GAS infections. While some of these are well established (e.g., varicella, child birth, lacerations), most host factors that control susceptibility and outcome are not understood. Lastly, as remarkable as the recent appearance of these cases has been, the continued persistence of severe invasive GAS infections is enigmatic and equally intriguing.
1. Cockerill FR, MacDonald KL, Thompson RL, et al. An outbreak of invasive Group A streptococcal disease associated with high carriage rates of the invasive clone among school-aged children. JAMA 1997; 277 (1):38–43.
2. Auerbach SB, Schwartz B, Williams D, et al. Outbreak of invasive group A streptococcal infections in a nursing home. Lessons on prevention and control. Arch Intern Med 1992; 152 (5):1017–22.
3. Hohenboken JJ, Anderson F, Kaplan EL. Invasive group A streptococcal (GAS) serotype M-1 outbreak in a long-term care facility (LTCF) with mortality. In: Program & Abstracts of the Interscience Conference on Antimicrobial Agents and Chemotherapy, Orlando, FL. Abstract# J189.1994.
4. Centers for Disease Control. Nosocomial group A streptococcal infections associated with asymptomatic health-care workers – Maryland and California, 1977. MMWR Morb Mortal Wkly Rep 1999;48:163–6.
5. The Working Group on Prevention of Invasive Group A Streptococcal Infections. The working group on prevention of invasive group A streptococcal infections. Prevention of invasive group A streptococcal diseases among household contacts of case-patients: is prophylaxis warranted? JAMA 1998;279:1206–10.
6. Stevens DL, Tanner MH, Winship J, Swarts R, et al. Reappearance of scarlet fever toxin A among streptococci in the Rocky Mountain West: Severe group A streptococcal infections associated with a toxic shock-like syndrome. N Eng J Med 1989; 321 (1):1–7.
7. Stevens DL. Could nonsteroidal anti-inflammatory drugs (NSAIDs) enhance the progression of bacterial infections to toxic shock syndrome? Clin Infect Dis 1995; 21:977–80.
8. Demers B, Simor AE, Vellend H, et al. Severe invasive group A streptococcal infections in Ontario, Canada: 1987–1991. Clin Infect Dis 1993; 16:792–800.
9. Stegmayr B, Bjorck S, Holm S, et al. Septic shock induced by group A streptococcal infections: clinical and therapeutic aspects. Scand J Infect Dis 1992; 24:589–97.
10. Stevens DL, Bryant AE, Hackett SP, et al. Group A streptococcal bacteremia: The role of tumor necrosis factor in shock and organ failure. J Infect Dis 1996; 173:619–26.
11. Stevens DL. Invasive group A streptococcus infections. Clin Infect Dis 1992; 14:2–13.
12. Meleney FL. Hemolytic Streptococcus Gangrene. Arch Surg 1924; 9:317–64.
13. Dick GF, Dick GH. Experimental scarlet fever. JAMA 1923; 81:1166.
14. Dick GF, Dick GH. The etiology of scarlet fever. JAMA 1924; 82:301.
15. Dick GF, Dick GH. Therapeutic results with concentrated scarlet fever antitoxin. JAMA 1925; 84:803.
16. Dick GF, Dick GH. The prevention of scarlet fever. JAMA 1924; 83:84.
17. Stevens DL. Group A beta hemolytic streptococci: Virulence factors, pathogenesis, and spectrum of clinical infections. In: Stevens DL, Kaplan EL, editors. Streptococcal Infections: Clinical Aspects, Microbiology and Molecular Pathogenesis. New York: Oxford University Press, 2000: 19–36.
18. Stevens DL, Salmi DB, McIndoo ER, Bryant AE. Molecular epidemiology of nga and NAD glycohydrolase/ADP-ribosyltransferase activity among Streptococcus pyogenes causing streptococcal toxic shock syndrome. J Infect Dis 2000; 182:1117–28.
19. Stevens DL, Yan S, Bryant AE. Penicillin binding protein expression at different growth stages determines penicillin efficacy in vitro and in vivo: An explanation for the inoculum effect. J Infect Dis 1993; 167:1401–5.
20. Stevens DL, Bryant AE, Hackett SP. Antibiotic effects on bacterial viability, toxin production, and host response. Clin Infect Dis 1995; 20:S154–S157.
21. Hirata N, Hiramatsu K, Kishi K, et al. Pretreatment of mice with clindamycin improves survival of endotoxic shock by modulating the release of inflammatory cytokines. Antimicrob Agents Chemo 2001; 45:2638–42.
22. Stevens DL, Bryant-Gibbons AE, Bergstrom R, et al. The Eagle effect revisited: Efficacy of clindamycin, erythromycin, and penicillin in the treatment of streptococcal myositis. J Infect Dis 1988; 158:23–8.
23. Zimbelman J, Palmer A, Todd J. Improved outcome of clindamycin compared with beta-lactam antibiotic treatment for invasive Streptococcus pyogenes infection. Pediatr Infect Dis J 1999; 18:1096–100.
24. Norrby-Teglund A, Basma H, Andersson J, et al. Varying titres of neutralizing antibodies to streptococcal superantigens in different preparations of normal polyspecific immunoglobulin G (IVIG): implications for therapeutic efficacy. Clin Infect Dis 1998; 26:631–8.
25. Norrby-Teglund A, Stevens DL. Novel therapies in streptococcal toxic shock syndrome: attenuation of virulence factor expression and modulation of host response. Curr Opin Infect Dis 1998; 11:285–91.
26. Kaul R, McGeer A, Norrby-Teglund A, et al. Intravenous immunoglobulin therapy for streptococcal toxic shock syndrome - A comparative observational study. Clin Infect Dis 1999; 28:800–7.