Toxic Shock Syndrome Management: A tale of two patients

Introduction

Toxic shock syndrome (TSS) is a true resuscitationist's disease. It is potentially quite lethal, with many series of streptococcal toxic shock syndrome reporting mortality in the range of 30-50%. However, recent observational studies suggest that treatment with modern critical care, toxin-suppressive antibiotics, and IVIG may reduce the mortality to 10% (Linner 2014).

After last week's post on diagnosis, this post explores management. Treatment primarily consists of high-quality sepsis care, but there are some important points that are unique to TSS. This is illustrated in two examples.

Patient #1

Ten years ago, a 30-year-old woman presented to Genius General Hospitalwith a diffuse erythematous rash, diarrhea, nausea, and vomiting. She had developed flu-like symptoms and spent some days at home trying to rehydrate herself. When she could no longer get out of bed, she presented to the emergency department and was found to be in septic shock with renal failure. She was admitted to the ICU, resuscitated with fluid and vasopressors, and treated with broad-spectrum antibiotics including clindamycin. Intravenous immunoglobulin (IVIG) was not given because of the lack of proven mortality benefit in the literature. Her initial ICU course was marked by anuria and a moderate requirement for vasopressors (about 10-15 mcg/min norepinephrine). Imaging studies questioned an ovarian abscess, perhaps related to her intrauterine device (IUD). Due to failure to improve, she was taken to the operating room for surgical exploration. Following intubation she developed pulseless electrical activity and could not be resuscitated. Autopsy revealed an ovarian abscess containing group A streptococcus.

In reflecting on this case, perhaps the most notable feature is how the patient really didn't look that sick. She was mentating adequately without significant distress for several hours in the ICU. This seems to be a common error in treating young TSS patients: they may not appear to be very sick and thus don't receive sufficiently aggressive intervention. A very similar case was recently reported by Cho 2013.

Patient #2

More recently, a 35-year-old woman presented to an outside hospital with diffuse erythematous rash, hypotension, renal failure (creatinine 2 mg/dL), and severe thigh pain. She remained hypotensive despite four liters of fluid resuscitation. Upon transfer to the Genius General Hospital ICU, she was found to have a mean arterial pressure (MAP) in the low 50s with tachycardia, moderate tachypnea, and a lactate of 10 mM. Peripheral norepinephrine was immediately started to stabilize her MAP, requiring 15 mcg/min to achieve a MAP>65 mm. She was started on clindamycin, linezolid, piperacillin-tazobactam, and IVIG (one gram/kg). Although she wasn't in severe distress, she was intubated using 1.4 mg/kg Rocuronium and 0.4 mg/kg ketamine. Despite fluid resuscitation, following intubation her vasopressor requirement increased to norepinephrine 40 mcg/min and vasopressin 0.04 units/min. Stress dose hydrocortisone was started. CT scan of the thigh revealed focal soft tissue swelling. Surgical exploration was performed to evaluate for necrotizing fasciitis or any drainable abscess, however neither was found.

Due to persistent requirement for high doses of vasopressors and lack of a drainable focus, the critical care and infectious disease teams decided to trial high-dose IVIG (an additional 2 grams/kg, about two liters of fluid). After this dose of IVIG, her levophed infusion was titrated from 40 mcg/min to 4 mcg/min. The following morning she was extubated and vasopressors were weaned off. Meanwhile her renal function had normalized. She made a complete recovery and returned home a few days later.

Prompt hemodynamic and airway control

These patients are often much younger and fitter than most patients with septic shock. They often have enormous physiologic reserve, and may look deceptively OK. This may cause delays in providing adequate supportive therapies such as vasopressors and mechanical ventilation.

A previous post discussed the concept of immediate support of the mean arterial pressure (MAP) with norepinephrine, rather than waiting to see if fluid resuscitation would improve the blood pressure. Patients with severe TSS will often require rapid escalation to very aggressive vasoactive support in addition to volume resuscitation (e.g., norepineprhine, vasopressin, and stress-dose hydrocortisone). Renal failure is very common in TSS, so a perfusing MAP should be aggressively pursued.

Early pre-emptive airway control may be advisable for several reasons. Ventilatory support may relieve the patient's work of breathing, allowing blood to be directed to vital organs rather than the diaphragm. TSS patients often require transportation to the CT scan and/or operating room, and it may be safer to control the airway prior to transportation. Intubation should be performed after hemodynamic optimization, with every possible measure taken to avoid hemodynamic collapse (for a discussion of intubating the shocked patient see this lecture by Scott Weingart).

Rapid sequence intubation and procedurization may provide a logistically efficient approach to achieve hemodynamic and airway control rapidly. More on this here.

Clinicians should remain alert to the possibility of a septic cardiomyopathy emerging after the initial hyperdynamic phase of septic shock. Some toxins have been shown to have direct myocardial toxicity (Murray 2005). Serial bedside echocardiography will identify this problem, which may respond well to inotropic support such as dobutamine.

Toxin-suppressive antibiotics

The primary problem in TSS is often not the presence of bacteria in the tissues, but rather toxin secretion causing massive cytokine release. Therefore, antibiotic therapy focuses on protein-synthesis inhibiting drugs which act immediately to shut off toxin synthesis.

Clindamycin is the most widely recommended antibiotic for toxin suppression. Clindamycin has activity against >99% of Group A streptococci as well as some activity against staphylcooccci including some MRSA. Clindamycin acts on the 50s ribosomal subunit to inhibit protein synthesis, thereby stopping toxin synthesis. Data supporting the use of clindamycin consists of in vitro studies as well as retrospective clinical data (Schlievert 1984; Linner 2014).

TSS may occur due to staph infection including MRSA, or rarely Group A streptococci resistant to clindamycin. Therefore, adding a second antibiotic is advisable. Linezolid may be a logical choice here, because it has excellent coverage for gram positive pathogens and acts similarly to stop toxin synthesis by inhibiting the 50s ribosomal subunit. Although linezolid and clindamycin both act on the 50s ribosomal subunit, the two agents combined appear to be more effective than either alone (Sweeney 2003, Coyle 2003). Linezolid has been shown to suppress toxin production by both staph aureus and Group A streptococci, with clinical success against TSS in case reports (Stevens 2006; Coyle 2003; Shupp 2013).

Antibiotic effect on the growth of Staph Aureus (top) and the production of toxic shock syndrome toxin (TSST-1, bottom). Antibiotics were added to bacteria four hours after the initiation of cultures. Nafcillin, clindamycin, and linezolid were equally effective in reducing bacterial concentration (top figure). However, only linezolid and clindamycin were able to reduce toxin secretion (bottom figure; Stevens 2006).

Beta-lactams and vancomycin act by increasing bacterial wall permeability and causing bacterial lysis. In vitro, beta-lactams may actually increasetoxin secretion. Data supporting the use of beta-lactams for TSS is questionable, with evidence of treatment failure in animal models and potential harm due to increased toxin secretion (Stevens 2007).

Intravenous immunoglobulin (IVIG)

Toxic shock occurs in a minority of people who lack protective antibodies. Pooled immunoglobulin (IVIG) provides such antibodies, and in theory could be a powerful treatment. This remains a controversial topic, primarily because IVIG is fairly expensive. What does the data show?

Darenberg 2003 performed a multicenter, randomized, placebo-controlled trial of IVIG in strep TSS. Although this was a well-designed study, it was terminated after recruiting 21 patients due to slow recruitment. Treatment with IVIG reduced 28-day mortality from 36% to 10%, but this was not statistically significant (p = 0.3). Examining SOFA scores, it was observed that patients receiving IVIG improved whereas patients who received placebo did not improve over time:

Several observational studies have found a mortality benefit from IVIG (table below; Curtis 2014). Unfortunately these studies are subject to numerous confounders. For example its possible that IVIG use is a marker for a more aggressive resuscitation strategy overall.

Recently an additional prospective observational study found that IVIG and clindamycin use both correlated with survival, even after multiple logistic regression to correct for confounding factors (Linner 2014):

IVIG is generally regarded as safe. Anaphylaxis may occur in some patients with IgA deficiency, but this risk appears to be exceptionally low. Numerous RCTs have been performed investigating the utility of IVIG for septic shock, and analysis of these studies has not revealed a safety concern with IVIG (Cochrane Database 2013). Given that IVIG is safe, and potentially very helpful, the benefit likely outweighs the risk.

Most authors recommend a treatment regimen for TSS based on Darenberg 2003 (1 g/kg IVIG initially, followed by 0.5 g/kg on days 2-3). For patients refractory to this therapy, it is reasonable to escalate to higher doses (e.g. 2 g/kg/day, based on data from Norrby-Teglund 2005 discussed further below). The optimal dose remains unclear.

There are no clinical trials of IVIG in staph TSS, although IVIG has appeared to be effective in animal models and case reports. This seems reasonable pending further study. There is some evidence that staph toxin may be less responsive to IVIG (Darenberg 2004). For a patient with staph TSS who fails to respond to 1 g/kg IVIG, it may be reasonable to trial a higher dose.

Definitive source control

Source control is a cornerstone of sepsis management in general. This is especially critical if the focus of infection is continuing to secrete toxin. Source control may be challenging because TSS may result from an unimpressive nidus of infection (e.g., a small abscess). There should be a low threshold to obtain imaging studies, although normal imaging does not exclude necrotizing infection (Anderson 2014). Surgeons should be involved early with consideration to debride or drain any possible infectious focus. Surgical wounds should be considered potentially infected even if they appear benign. The presence of a retained tampon, sinus packing, or any infected foreign material must be excluded.

However, for patients with necrotizing fasciitis and strep TSS, immediate radical excision of involved tissue might be inadvisable. Norrby-Teguland 2005published a case series of seven patients with group A streptococcal necrotizing fasciitis (six of whom also had TSS). They took a conservative surgical approach with intensive medical therapy including high-dose IVIG (2 grams/kg on day one, and again on days 2-3 if needed). Medical management appeared to reduce the area of involved skin, allowing less extensive debridement (figure above). Low 2013 reviewed the literature and pointed out that there is no evidence to support the dogma that group A streptococcal necrotizing fasciitis should be immediately derided. This remains controversial. Surgeons, intensivists, and infectious disease specialists should be involved in these cases early, with treatment decisions based upon close collaboration and meticulous observation over time.

Conclusions

TSS is unusual in its ability to cause severe septic shock in previously normal young people. Given that patients often have high physiologic reserve, they may initially look deceptively OK, masking the severity of their illness. This often causes delays in instituting sufficiently aggressive management. In addition to aggressive supportive care, specific therapies including IVIG and toxin-suppressive antibiotics may be very important. The following checklist includes some interventions which should be considered when treating a patient with TSS.

Hemodynamic support
- Early norepinephrine to establish a MAP>65mm
- Fluid resuscitation
- Low threshold to add vasopressin and stress-dose hydrocortisone if not responding well to norepinephrine
- Serial echocardiography to identify occurance of septic cardiomyopathy
Respiratory support
- Consider pre-emptive intubation (e.g., for severe shock or respiratory distress). Meticulous attention required to peri-intubation hemodynamics.
Dual toxin-supressive antibiotic therapy
- Clindamycin 900 mg IV q8 hours first dose STAT
- Linezolid 600 mg IV q12 hours first dose STAT
Intravenous immunoglobulin (IVIG)
- Standard course: 1 g/kg IVIG on day #1, then 0.5 g/kg IVIG on days #2 and #3
- Consider higher dose IVIG (2 g/kg) for refractory shock or necrotizing fasciitis
Source control
- Early surgical consultation for debridement/drainage of any focus of infection
- Definitive imaging if needed to search for a focus of infection (e.g. CT scan, pelvic ultrasound, etc.)
- Exclude retained tampon, nasal packing, or foreign body