Severe Group A and Group B Streptococcus Diseases at a Pediatric ICU: Are they Still Sensitive to the Penicillins?

BACKGROUND: Group A β-hemolytic Streptococcus (GAS) and Group B streptococcus (GBS) are two common pathogens that are associated with many diseases in children. Severe infections as a result of these two streptococci are albeit uncommon but associated with high mortality and morbidity, and often necessitate intensive care support. This paper aims to review the mortality and morbidity of severe infection associated with GAS and GBS isolations at a Pediatric Intensive Care Unit (PICU).
METHODS: All children admitted to PICU of a teaching hospital between October 2002 and May 2018 with laboratory-proven GAS and GBS isolations were included.
RESULTS: There were 19 patients (0.7% PICU admissions) with streptococcal isolations (GAS, n=11 and GBS, n=8). Comparing to GAS, GBS affected infants were younger (median age 0.13 versus 5.47 years, 95% CI, 1.7-8.5, p=0.0003), and cerebrospinal fluids more likely positive (p = 0.0181). All GAS and GBS were sensitive to penicillin (CLSI: MICs 0.06 - 2.0 μg/mL), with the majority of GAS sensitive to clindamycin and erythromycin, and half of the GBS resistant to clindamycin and erythromycin. Co-infections were prevalent, but viruses were only isolated with GAS (p=0.024). Isolation of GAS and GBS was associated with nearly 40% mortality and high rates of mechanical ventilation and inotropic supports. All non-survivors had high mortality (PIM2) and sepsis scores.
CONCLUSIONS: Severe GAS and GBS are rare but associated with high mortality and rates of mechanical ventilation and inotropic supports in PICU. The streptococci are invariably sensitive to penicillin. The high PIM2 and Sepsis scores suggest that prompt recognition of sepsis and the timely judicious institution of antibiotics and intensive care support may be life-saving for these devastating infections. Copyright© Bentham Science Publishers; For any queries, please email at epub@benthamscience.net.

Introduction

The streptococci are Gram-positive non-motile and non-spore-forming bacteria which grow in chains [1]. Group A β-hemolytic Streptococcus (GAS) and Group B Streptococcus (GBS) are the two common streptococcal pathogens that are associated with many diseases in children [2, 3].

GAS produces extracellular enzymes and toxins with haemolytic properties [1, 4]. There were 1.78 million new cases of GAS infections each year [5]. It has been reported that acute GAS infections predominantly affect children 5 years to 15 years of age [6], and Streptococcus pyogenes (S. pyogenes) has been considered as the most important pathogen in GAS infections as it is responsible for causing many important human diseases, especially towards children [4]. It has been reported that the carrier rate of S. pyogenes in the throat of asymptomatic school children is up to 15-20% in some studies [6]. However, with a high carrier rate of GAS in children, it is sometimes difficult to determine if a child presenting with fever and sore throat having a positive GAS culture is due to a concurrent viral infection in a carrier or genuine GAS pharyngitis [1]. Infections with GAS typically begin in the throat or skin [6]. The spectrum of infections ranges from mild pharyngitis (strep throat) and localized skin infection (impetigo) to moderate-severe manifestations in the form of scarlet fever, pneumonia, bacteraemia, erysipelas, cellulitis, and to life-threatening conditions such as necrotizing fasciitis and toxic shock syndrome which requires intensive care supports [7, 8]. Clinical presentations of streptococcal infections vary according to the virulence factors of the bacteria and individual host response [7], and the severe streptococcal infections can be classified into 5 groups, namely, streptococcal toxic shock syndrome, invasive infections, scarlet fever, non-invasive infections (throat and skin infections), and non-suppurative sequelae according to a working group formed by an informal group of microbiologists, clinicians and epidemiologists in 1993 [3, 7]. GAS-associated late immune-mediated conditions include rheumatic fever, post-streptococcal glomerulonephritis, arthritis and Paediatric Autoimmune Neuropsychiatric Disorders 
Associated with Streptococcal Infections (PANDAS) also contributing to the significant mortality and morbidity [9].

GBS is a common microbe colonizing the maternal gastrointestinal and genital tract [10]. The organism can spread from the mother to the neonate during vaginal delivery. Most neonatal infections caused by GBS occur within the first 6 days of life [11-14]. While early-onset GBS infection (from birth to 6 days) often presents as sepsis, pneumonia, late-onset GBS infection (from 7 days to 89 days) more commonly presents as bacteremia, meningitis, and, less commonly, other focal infections such as cellulitis, lymphadenitis, and bone/joint infection, contributing to the need of intensive care supports [2, 11-15].

Children are prone to infections, especially towards these two group streptococci (5, 10-13). Severe infections, with life-threatening complications as a result of these two streptococci are albeit uncommon but associated with high mortality and morbidity, and often necessitate intensive care support [8, 16]. However, epidemiology of these severe infections in pediatric intensive care settings is seldom reported. This paper aims to review the mortality and morbidity of severe infection associated with GAS and GBS isolations at a Pediatric Intensive Care Unit (PICU).

Methods

Children admitted to PICU of a university-affiliated teaching hospital [Prince of Wales Hospital (PWH)] between October 2002 and April 2018 with laboratory-proven GAS and GBS isolations were audited. PWH provides tertiary PICU service for children aged up to 14 years in the Eastern New Territories of Hong Kong with a catchment population of over 1.1 million (approximately 25% were children less than 12 years of age) [17]. Children with streptococcal isolates were identified by using the principal author’s database that registered every PICU admission. Mortality risk was measured by the Paediatric Index of Mortality version 2 score (PIM 2: http://www.sfar.org/scores2/pim22.html) [18]. Sepsis-1, Sepsis-2 and Sepsis-3 definitions were retrospectively defined for the non-survivors [19, 20]. The bacteria were identified using conventional diagnostic methodology from specimens including blood cultures, Cerebrospinal Fluids (CSFs), pleural aspirates, tracheal aspirates (via endotracheal tube in intubated patients), sputum, urine and deep wound swabs (e.g. retropharyngeal abscess, throat, peritoneum, eyes, umbilicus). The diagnostic procedures were performed in the same microbiology laboratory with standardized procedures and equipment. The antibiotic susceptibilities of S. pyogenes and S. agalactiae against penicillin, erythromycin, and clindamycin performed by disc testing and the Minimum Inhibitory Concentrations (MICs) of penicillin was determined using microbroth dilution method and interpreted according to criteria set by the Clinical Laboratory Standards Institute.

Demographic data, mortality, length of PICU stay, hospital stay (till discharge or death), complications, specimen site and co-infection were compared with chi-square test (for numerical values), Fisher’s exact test (valid for all sample sizes, especially when sample sizes are small), or Mann–Whitney U test (for non-parametric data). Univariate and multivariate models were used for independent predictors and various confounding factors. All comparisons were made two-tailed, and p values less than 0.05 were considered statistically significant. Ethics approval for this audit was obtained from the Joint Chinese University of Hong Kong – New Territories East Cluster Clinical Research Ethics Committee.

Results

As of April 2018, S. pyogenes (GAS, n = 11) and S. agalactiae (GBS, n = 8) were isolated among 2660 PICU admissions, giving an incidence of 0.4% and 0.3%, respectively (Table ). Comparing the two microbes, CSFs was more likely positive for GBS than GAS and (p = 0.0181). All GAS and GBS were sensitive to penicillin (CLSI: MICs 0.06 – 2.0 μg/mL), with the majority of GAS sensitive to clindamycin and erythromycin, and more than half, 75% of the GBS resistant to clindamycin and erythromycin.

Bacterial, viral and fungal coinfections were found in four (methicillin-resistant S. aureus, n = 1; S. aureus, n = 2; Pseudomonas sp., n = 1; Acinetobacter sp., n = 1; Escherichia coli, n = 1; Conebacter sp., n = 1; Enterobacter sp., n = 1, S. viridans, n = 1, Rothia mucilaginosa, n = 1), six (respiratory syncytial virus, n = 2; influenza A, n = 2; parainfluenza, n = 1; Epstein–Barr virus, n = 1; Herpes simplex virus, n = 1; varicella zoster virus, n = 1) and one (Candida glabrata, n = 1; Candida albicans, n = 2) patients respectively in GAS; and only bacterial coinfection was found in three patients (S. aureus, n = 1; Pseudomonas sp., n = 1; Klebsiella sp., n = 1; Micrococcus sp., n=1) in GBS. Viral coinfection was significantly associated with GAS isolation than GBS (p=0.0237) (Table ).

Isolation of GAS and GBS was associated with a 37% mortality and high rates of need for mechanical ventilatory and inotropic supports in the PICU setting; GAS infections were associated with PICU mortality of 36% (n = 4) and high rates of need for mechanical ventilatory and inotropic supports, 64% and 55% respectively, whereas GBS infections were associated with PICU mortality of 38% (n = 3) and high rate (50%) of need for ventilatory support only. GBS infection significantly lengthens the PICU stay of these patients. Non-survival cases of GAS and GBS were summarized in Tables and , respectively. No significant difference was found between the survived versus non survived groups. All non-survivors had high mortality (PIM2) and sepsis scores (Tables 2 & 3: Sepsis-1, Sepsis-2 and Sepsis-3 according to definition in methodology) [18-20].

Discussion

GAS and GBS in PICU Setting

Group A β-hemolytic streptococcus (GAS) and Group B Streptococcus (GBS) are common streptococcal pathogens that are associated with many diseases in children [6, 15]. Severe streptococcal infections are rare but a major factor of childhood mortality. The GAS and GBS infect different age groups of children and result in different spectra of invasive diseases, such multiorgan dysfunction syndrome, septicemia, meningitis or pneumonia, resulting in intensive care treatment. Severe streptococcal infections due to the two pathogens in the PICU setting are fortunately relatively rare. In our series, these two organisms are only isolated in 0.7% of PICU admissions over a 15-year period. However, isolation of GAS and GBS in PICU setting was associated with an alarming 37% mortality and high rates of need for mechanical ventilatory and inotropic supports.

Non-survivors with GAS Infections

There were 4 PICU non-survivors with GAS infections, with a mean age of the 8.01 years (Table ). The average Length Of Stay (LOS) in PICU (till death) was 0.97 day. Blood cultures were positive for GAS in 75% cases and all GAS isolations were sensitive to penicillin. All non-survivors were at least co-infected with one other virus or bacteria and presented with sepsis at the time admission, with all fulfilled the criteria of Sepsis-1, -2 and -3 [19, 20]. It is difficult to attributed non-survival solely to the streptococci due to the presence of co-infections. The significance of co-infections cannot be evaluated due to small sample size. Immunity of these patients may be compromised during severe sepsis leading to co-infections by multiple organisms in various organ sites. Broad-spectrum antibiotic coverage is called for critically ill patients. The subsequent antibiotic regime must then be guided by culture results. Most of the non-survivors were admitted through the emergency department (n = 3) and only one of the three non-survivor was given antibiotic prior to the admission to PICU. Clinical feature suggestive of scarlet fever was presented in one of the non-survivors.

LOS, Length of stay; PIM2, Paediatric Index of Mortality version 2 score; [18] A&E, Accident & Emergency Department, HSV, Herpes Simplex virus; MRSA, methicillin-resistant S. aureus; RSV, respiratory syncytial virus.

The mortality risk of non-survivors was accessed by the PIM2 score. With the admission status of having sepsis while low rate (25%) of antibiotic usage prior PICU admission, the average PIM2 score of the non-survivor was 78%, showing that the sepsis assessments and antibiotic usage prior PICU admission may have impacts on the mortality of the GAS infection in PICU.

GAS is generally sensitive to beta-lactams group of antibiotics [3, 4, 19, 21]. Failure of treatment with penicillin has been attributed to other local commensal organisms producing β-lactamase, or failure to achieve adequate tissue levels in the pharynx. A ten-day course of penicillin or a narrow spectrum cephalosporin in patients with penicillin allergy is recommended for the treatment of GAS pharyngitis and scarlet fever [21]. Amoxicillin is often preferred over penicillin V because of the better taste of the suspension and its availability as chewable tablets [1]. Effective treatment of GAS infection is important for the control of acute infection and the prevention of late immunological manifestations of rheumatic heart disease, glomerulonephritis and PANDAS. Despite the availability of antibiotics, GAS remains an important bacterial pathogen causing a wide variety of diseases in children. Early recognition of symptoms and signs of invasive infections and serious toxin-mediated conditions prevent fatality [8]. Late sequelae, mortality and morbidity of severe GAS may be avoided.

Non-survivors with GBS Infections

There were 3 PICU non-survivors with GBS infections, with a mean age of the 0.84 years (Table ). The average Length Of Stay (LOS) in PICU (till death) was 67.3 days. Blood cultures were also likely to be positive for GBS and all GBS isolates were sensitive to penicillin. All non-survivors were at least co-infected with one bacteria, with no viral co-infection) and presented with sepsis at the time admission, with all fulfilled the criteria of Sepsis-1, -2 and -3 [19, 20]. Two of the three non-survivors were admitted through the emergency department, and one of the two non-survivors was given antibiotics in the emergency department prior to admission to PICU. The mortality risk of non-survivors was assessed by the PIM2 score [18]. With the admission status of having sepsis while low rate of antibiotic usage prior PICU admission, the average PIM2 score of the non-survivor was 61%, showing that the sepsis assessments and antibiotic usage prior PICU admission may also have impacts on the mortality of the GBS infection in PICU.

Limitations

Streptococcal isolation is not prevalent in this study. Most PICUs in Hong Kong are small units, and severe infections are relatively less common among children in this affluent city. In severe infections, empirical antibiotic usage prior to PICU admission may lead to negative cultures in some of these septic patients.

Conclusions

In summary, among the non-survived cases, PIM2 reflects high mortality risk [18]. In addition, these cases fulfill various sepsis criteria. The two streptococci are invariably sensitive to penicillin in our locality. All these point to the timely use of antibiotics in children present with severe sepsis. Primary care physicians might refrain from prescribing antibiotics in seemingly trivial infections with or without an obvious focus, such as pharyngitis and occult bacteremia. The non-survivors of these streptococcal infections, often previously healthy children, serve to remind us to be vigilant on sepsis, and timely appropriate use of antibiotics may be life-saving and justifiable. As suggested by the sepsis campaign, physicians and the community must be more aware of the devastating sequelae of sepsis, a disease that is readily amenable to timely antibiotic and supportive treatment.

Table 1

Patient demographics of streptococcal infections in PICU.

-	GAS (n = 11)	GBS (n = 8)	P
Males (%)	5(45%)	5(63%)	0.46
Median (IQR) age	5.47	0.13	0.0003*
-	(1.83-8.62)	(0.086-0.84)	-
Blood culture	4	7	0.0587
CSF	0	4	0.0181*
Lung tissue/pleural fluid/tracheal aspirate/sputum	5	1	0.13
Miscellaneous^	6	1	0.061
Coinfection identified	8	4	0.16
Viral coinfections	6	0	0.0124*
Bacterial coinfections	4	5	0.37
Fungal coinfections	1	0	0.38
Penicillin	-	-	-
Sensitive	11	8	-
Resistant	0	0	-
Clindamycin	-	-	-
Sensitive	8	4	-
Resistant	3	4	-
Erythromycin	-	-	-
Sensitive	8	4	-
Resistant	3	4	-
Mechanical ventilation	7	4	0.55
Inotropes	6	1	0.12
Corticosteroids	4	2	0.84
Median (IQR) PICU stay (days)	2 (0.77-4.75)	5.5 (3.25-51.5)	0.035*
Death	4	3	>0.99
Median (IQR) hospital stay (days)	8 (4-19)	34 (10.25-61.75)	0.17

^: wound, eye, throat, umbilical and peritoneal swabs. GAS, Group A streptococcus; GBS, Group B streptococcus; PICU, paediatric intensive care unit; IQR, interquartile range.

Table 2

Nonsurvivors in the PICU with GAS infection.

Sex	Age (Years)	LOS in PICU (Days)	Infection Site(s)	Antibiotic Sensitivity	Co-infection(s)	Scarlet Fever	PIM2 (%)	A&E Admission	Antibiotics Prior PICU	Sepsis-1*	Sepsis-2*	Sepsis-3*
M	10.23	0.83	Blood culture and tracheal aspirate	Sensitive: penicillin, clindamycin, erythromycin	Influenza A	No	90.0	Yes	No	Yes	Yes	Yes
F	6.88	0.042	Skin and eye swabs, pleural fluid	Sensitive: penicillin, clindamycin, erythromycin	Acinetobacter species, Candida glabrata, Candida lusitaniae, Corynebacter species, Diptheroids, E. coli (on wounds), Enterobacter species, HSV, MRSA, pseudonomas aeruginosa, RSV, Staphylococcus aureus	No	96.6	No	No	Yes	Yes	Yes
F	4.26	3	Blood culture	Sensitive: penicillin, clindamycin, erythromycin	Staphylococcus aureus, Gram positive Cocci	No	28.1	Yes	No	Yes	Yes	Yes
F	10.68	0	Spleen and lung tissue, blood culture	Sensitive: penicillin, clindamycin, erythromycin	Streptococcus viridans, Rothia mucilaginosa, Actinomyces species	Yes	99.2	Yes	Yes	Yes	Yes	Yes

*Sepsis-1 defines as at least two of the following Systemic Inflammatory Response Syndrome (SIRS) clinical criteria are present: heart rate faster than 90 beats per min; respiratory rate faster than 20 breaths per min; leukocyte count greater than 12,000/μL or less than 4,000/μL; temperature less than 36°C or greater than 38°C, within the first 24 h of admission to PICU; Sepsis-2 defines as at least two of the SIRS criteria with confirmed infections; Sepsis-3 defines as at least two of the three clinical criteria in the pre-screening by Sequential Organ Failure Assessment (qSOFA) followed by at least two acute change of the complete SOFA score with confirmed infection [19, 20].

Table 3

GBS infection: survivors and nonsurvivors in the PICU.

Sex	Age (Year(s))	LOS in PICU (Day(s))	Death	Infection Site(s)	Antibiotic Sensitivity	Co-infection(s)	Scarlet Fever	PIM2 (%)	A&E Admission	Antibiotics Prior PICU	Sepsis-1	Sepsis-2	Sepsis-3
M	1.03	1	Yes	Blood culture	Sensitive: penicillinResistant: clindamycin, erythromycin	Staphylococcus aureus	No	76.9	Yes	No	Yes	Yes	Yes
M	1.45	135	Yes	Lung tissue	Sensitive: penicillinResistant: clindamycin, erythromycin	Pseudomonas species	No	50.1	No	No	Yes	Yes	Yes
M	0.05	66	Yes	CSF, blood culture and umbilical swab	Sensitive: penicillinResistant: clindamycin, erythromycin	Klebsiella species	No	55.5	Yes	Yes	Yes	Yes	Yes
M	0.08	3	No	Blood culture	Sensitive: penicillin, cotrimoxazole, erythromycin	NA	No	0.8%	Yes	No	Yes	Yes	No
M	0.28	8	No	CSF, blood culture	Sensitive: penicillin, clindamycin, cotrimoxazole	NA	No	2.6%	No	Yes	Yes	Yes	Yes
F	0.12	7	No	CSF, blood culture	Sensitive: penicillinResistant: clindamycin, erythromycin	Micrococcus species	No	11.1%	Yes	Yes	Yes	Yes	Yes
F	0.13	4	No	Blood culture	Sensitive: penicillinResistant: clindamycin, erythromycin	NA	No	5.8%	Yes	No	Yes	Yes	No
F	0.10	4	No	Blood culture	Sensitive: penicillinResistant: clindamycin, erythromycin	NA	No	2.6%	Yes	Yes	Yes	Yes	Yes

*Sepsis-1 defines as at least two of the following Systemic Inflammatory Response Syndrome (SIRS) clinical criteria are present: heart rate faster than 90 beats per min; respiratory rate faster than 20 breaths per min; leukocyte count greater than 12,000/μL or less than 4,000/μL; temperature less than 36°C or greater than 38°C, within the first 24 h of admission to PICU; Sepsis-2 defines as at least two of the SIRS criteria with confirmed infections; Sepsis-3 defines as at least two of the three clinical criteria in the pre-screening by Sequential Organ Failure Assessment (qSOFA) followed by at least two acute change of the complete SOFA score with confirmed infection [19, 20].