de intensivist

Abstract
A 42-year-old male patient with a history of type 2 diabetes mellitus presented to the emergency department with haemodynamic and respiratory instability, resulting in an in-hospital cardiac arrest. The underlying cause was toxic shock syndrome caused by a necrotising diabetic foot infection with a beta-haemolytic group A Streptococcus of emm-type 89. This case report describes the clinical course, difficulty of diagnosis, pathophysiology and treatment of a patient with streptococcal toxic shock syndrome. The genotype of the group A Streptococcus determined by emm-typing and next-generation sequencing will be discussed in respect to the current epidemiological situation in the Netherlands.

Case
A 42-year-old male patient with a history of diabetes mellitus type 2 presented to the emergency department with haemodynamic and respiratory instability and signs of shock. The patient attended the surgical outpatient clinic earlier that day with an infected necrotising toe, for which conservative treatment was started with oral clindamycin and ciprofloxacin. At presentation the patient had a fever and sinus tachycardia of 137 beats/min with a blood pressure of 96/47 mmHg. The patient was in respiratory distress. The breathing frequency was 44/min with a saturation of 96% with 5 litres of oxygen. Within a few hours his condition deteriorated rapidly, followed by an acute cardiac arrest with pulseless electrical activity based on distributive shock. Other causes of cardiac arrest were excluded. No relevant electrolyte disorders were present, and a chest X-ray showed no signs of pneumothorax. Quick-look ultrasound did not reveal pericardial effusion or dilation of the right ventricle. There was no suspicion of intoxication. After two blocks of cardiopulmonary resuscitation (CPR), return of spontaneous circulation (ROSC) was achieved. The patient was intubated and admitted to the intensive care unit (ICU), where he suffered another cardiac arrest. ROSC was obtained shortly after initiation of CPR.

Before the first cardiac arrest, a markedly elevated C-reactive protein (400 mg/l) and leukocyte count (29*10⁹/l) were noted, accompanied by high anion gap metabolic acidosis, elevated lactate (5 mmol/l) and acute kidney injury (KDIGO AKI stage 3: Serum creatinine was 59 μmol/l during steady state, 624 μmol/l at presentation).

The patient was admitted to the ICU with the presumptive diagnosis of sepsis from the infected toe, for which he was treated with piperacillin/tazobactam and gentamicin. The second digit of the left foot showed dry necrosis and only mild redness of the forefoot. The left lower leg showed an erythematous lesion with an intersection of 5 cm. The infected toe was amputated. Peroperatively, well-circulated tissue with no signs of major infection was seen. Three days later wound cultures showed sporadic growth of a beta-haemolytic group A Streptococcus (GAS) and Staphylococcus aureus. Toxic shock syndrome was therefore added to the differential diagnosis.

Diagnostics
As shown in table 1, the diagnostic workup did not initially clarify the source of the sepsis. The urine sediment was contaminated, but the patient did not have urinary tract symptoms and CT showed no signs of pyelonephritis. Because of bad teeth, endocarditis was in the differential diagnosis. A transesophageal echocardiography (TEE) showed no valvular vegetations. The patient only had one minor Duke’s criterion which was fever. A CT cerebrum showed no dental focus for infection.

To help evaluate the plausibility of a streptococcal toxic shock syndrome, anti-streptolysin O and anti-DNAse B tests were performed. The results are shown in table 2. DNAse B and streptolysin O are enzymes produced by the majority of GAS. Anti-streptolysin O and anti-DNAse B were tested jointly, reducing the possibility of false negative outcomes.

Further evaluation of the isolated strains was performed by the Dutch Reference Laboratory for Bacterial Meningitis (NRLBM). The GAS isolated in the wound culture was emm-type 89, which is one of the common genotypes in invasive GAS infections.[1,2] The isolated S. aureus was toxic shock syndrome toxin-1 gene (TSST-1) negative, which excluded S. aureus as the causative agent.

Discussion
The diagnostic process in this case was difficult and time consuming. Extensive diagnostic testing initially did not clarify the infectious source of the sepsis. Severe septic shock is typically caused by only a few pathogens. These include exotoxin-producing bacteria such as S. aureus and beta haemolytic Streptococci and endotoxin-producing Meningococci and Caphnocytophaga canimorsus, typically found in bite wound infections.[3]

After amputation, the infected necrotising toe did not visually appear to be the source of sepsis, as there was almost no involvement of the soft tissue. Wound cultures showed growth at day three after admission, thus toxic shock syndrome was first considered three days after admission when the patient was already recovering. It took 14 days to get the results from the anti-streptolysin O and anti-DNAse B antibody tests, emm-genotyping took weeks and whole genome sequencing took more than a month. Anti-streptolysin O and anti-DNAse B do not help in the acute setting and do therefore not lead to a different treating policy. They did, however, help clarify the source of the sepsis. A rapid streptococcal antigen test, mostly used for throat infections, could possibly be of help in the acute situation.[4] In our case, an antigen test was not performed and is not readily available in our institution’s laboratory.

GAS are typed by differences in the gene (emm) coding for a surface antigen called the M-protein; emm-typing can help to determine whether or not different GAS infections belong to the same infection cluster. Certain emm-types are associated with streptococcal toxic shock syndrome.[1,2] The dominant type in invasive GAS infections in Europe and Northern America is emm-1, while emm-89 is accountable for a substantial number of invasive GAS infections.[2] Since 1 July 2019 microbiology laboratories in the Netherlands are requested to send GAS isolates of patients with invasive GAS infections to the NLBRM for typing. Of the received isolates, 11% were typed as emm-89, ranking as the third-most prevalent emm-type.

The isolated GAS strain was analysed using whole-genome sequencing and harboured polymorphisms that are associated with increased production of the toxins NADase and streptolysin O.[5,6] These polymorphisms are also found in emm-1 types, the classical toxic shock syndrome genotype. Increased production of NADase and streptolysin O in GAS is associated with increased virulence.[7]

Despite the lack of soft tissue involvement at the infection site, findings in antibody testing and genome typing established that streptococcal toxic shock syndrome was the cause of the haemodynamic instability and cardiac arrest. In toxic shock syndrome, superantigens produced by GAS bind non-specifically to the T-cell receptor of the CD4+ lymphocyte and major histocompatibility complex II on the antigen-presenting cell. The superantigens bind on the constant part of both receptors, bypassing the binding sites that only facilitate specific binding.[8, 9] Mass activation of CD4+ lymphocytes leads to mass secretion of pro-inflammatory cytokines, which results in increased permeability of blood vessels and vasodilation. One of the diagnostic criteria of toxic shock syndrome is a macular erythematous rash.[10] However, in the majority of patients this symptom is not observed, as in our patient.[11]

The Dutch SWAB guideline recommends benzylpenicillin or cefuroxime in combination with adjunctive clindamycin for streptococcal toxic shock syndrome. Intravenous immunoglobulin G (IVIG) should be considered.[12] Clindamycin and IVIG were not administered. Clindamycin blocks ribosomal function and thus toxin production. IVIG increases plasma-neutralising activity against superantigens.[13] Because of the fulminant course of streptococcal toxic shock syndrome, mediated by massive CD4+ lymphocyte activation by superantigens, it is likely that IVIG and clindamycin should be administered as early as possible in the clinical course. When streptococcal toxic shock syndrome was first considered as a potential diagnosis, the patient’s need for haemodynamic support was declining (noradrenalin 0.13 µg/kg/min at day 3 vs 1.35 µg/kg/min at day 1). Amputation of the infection site already stopped superantigen production and the time window for early intervention had already passed.

The adjunctive use of clindamycin in streptococcal toxic shock syndrome is generally accepted due to observational studies which showed reduced mortality in patients treated with clindamycin.[14,15] The evidence for IVIG is, however, limited. One randomised controlled trial (RCT) has been conducted, which was ended preliminarily because of slow patient recruitment; 21 patients were enrolled and were allocated to adjunctive treatment with IVIG or placebo. The trial showed a 3.6-fold higher 28-day mortality rate in the control group, which was not statistically significant.[15] A meta-analysis which included the aforementioned RCT and four prospective non-randomised case-control studies showed a reduction in mortality with a relative risk of 0.46 (95% CI 0.26-0.83).[16] The use of corticosteroids for streptococcal toxic shock syndrome has not been studied in observational nor interventional studies.

The antibiotic treatment with piperacillin/tazobactam was continued for a total duration of ten days. The patient recovered slowly and was able to be discharged from the ICU after ten days. After four weeks he was admitted to a clinical rehabilitation facility. At discharge the patient was in need of haemodialysis, which could be discontinued after three months. The deterioration and slow recuperation of his kidney function was interpreted as a combination of severe acute tubule necrosis and diabetic nephropathy.

Conclusion
Toxic shock syndrome may rapidly progress leading to multiorgan failure and death. In our patient, initiation of antibiotics and amputation may have prevented a worse outcome. This case shows that even without major inflammation of the wound, superantigens produced by GAS can be life-threatening. Prompt antibiotic treatment and elimination of the source of infection are of vital concern. The isolated GAS fits the description of an emm-89 lineage which became emergent in Europe and Northern America.

Disclosures
All authors declare no conflict of interest. No funding or financial support was received. Written informed consent was obtained from the patient for the publication of this case report and the accompanying image.