Recurrent group B streptococcus infection in an extremely premature infant: as a preterm neonate, infant and toddler
- 1 Paediatric Immunology and Infectious Diseases, Great North Children's Hospital, Newcastle upon Tyne, UK
- 2 Translational and Clinical Research Institute, Newcastle University, Newcastle upon Tyne, UK
- 3 Neonatalogy, Newcastle Upon Tyne Hospitals NHS Foundation Trust, Newcastle upon Tyne, UK
- Correspondence to Ina Schim van der Loeff; ina.schim@nhs.net
Abstract
We report five discrete episodes of group B streptococcus (GBS) bacteraemia in an extremely premature infant, extending into early childhood. The first four episodes occurred during infancy despite appropriate treatment. Breastmilk was positive for group B streptococcal 16S DNA by polymerase chain reaction. The fifth episode occurred at 17 months of age, shortly after stopping antimicrobial prophylaxis.
Radiological investigations did not identify a focus for recurrence of GBS bacteraemia, and immunological investigations and targeted whole genome sequencing yielded only transient hypogammaglobulinaemia of infancy, which resolved.
This case highlights invasive GBS infection as a cause of infant morbidity. Premature infants are at particular risk of invasive as well as recurrent disease. GBS is typically a sensitive organism and each episode of GBS in our patient was effectively treated with penicillin. The role of breastmilk in recurrent GBS is controversial; in this case infant and mother isolated identical GBS serotypes and were concurrently treated with rifampicin.
Background
Infection with group B streptococcus (Streptococcus agalactiae, GBS) remains a leading cause for neonatal disease worldwide and in the UK, with significant associated morbidity and mortality (6.2% case fatality rate in the UK).1 Infection can be early-onset (EOD, <7 days of life) or late-onset (LOD, 7–89 days of life). EOD is associated with vertical transmission from pregnant mothers, 20% of whom are colonised with GBS.2 LOD also arises from nosocomial or community transmission.
GBS is a significant cause of perinatal disease associated with puerperal sepsis, stillbirth, neonatal sepsis, preterm birth and neonatal encephalopathy. Several initiatives have attempted to reduce the impact of neonatal morbidity and mortality of GBS infection, including risk factor-guided use of intrapartum antibiotic prophylaxis (IAP) in labouring women. These strategies have been adopted in UK guidance published by the Royal College of Obstetricians and Gynaecologists, and the National Institute for Health and Care Excellence. IAP strategies prevent some infants developing EOD through vertical transmission, though their use does not prevent LOD.3 Indeed, prospective national surveillance of GBS in the UK and ROI has demonstrated a rise in incidence in GBS infection despite introduction of IAP guidance, though the relative rise in EOD incidence is smaller than that of LOD.1
GBS infection recurs in 0.5%–3% of infants with LOD,4 5 yet 20% of healthy infants are colonised by age 2 months.6 Proposed mechanisms for recurrence include repeated exposure to nosocomial reservoirs, colonisation of the infant’s gastrointestinal tract or of maternal breastmilk, persistence of an infective focus such as an abscess, and inadequate treatment agents or duration. Prematurity is a risk factor for EOD, LOD and recurrent GBS infection.7 Pathogen virulence factors such as capsular serotype are also implicated, with invasive and recurrent GBS disease isolating serotype III and its hypervirulent clonal complex 17 (CC17) in around two-thirds of cases.5 6
We report a case of five discrete episodes of GBS infection in a female infant born at extreme prematurity, recurring despite appropriate antimicrobial treatment.
Case presentation and treatment
A female infant was born by breech vaginal delivery at 23+2 gestation with a birth weight of 580 g. Membranes ruptured 1 week prior to delivery, and antenatal steroids, magnesium sulphate and prophylactic erythromycin were administered. Maternal bacteriology was negative.
The infant was intubated and given surfactant after delivery. She was treated with 48 hours of intravenous gentamicin and benzylpenicillin, which were stopped following negative blood cultures. Feeding with expressed breastmilk (EBM) was started on day 2 of life and titrated with parenteral nutrition until full enteral feeding was established at day 9. From 7 days of life, she was treated for 10 days with a continuous vancomycin infusion due to Staphylococcus epidermidis bacteraemia.
GBS was first isolated from a blood culture at 14 days of age when she developed hyperglycaemia, temperature instability and an increased oxygen requirement. She was treated with 14 days of antibiotics: 2 days of ceftazidime and vancomycin, then amoxicillin (three times per day intravenous 50 mg/kg) for 5 days once blood culture results were known. Due to concerns about necrotising enterocolitis, antibiotics were changed to ceftazidime, vancomycin and metronidazole for a further 7 days. Feeding was stopped and she required parenteral nutrition for 7 days before returning to full EBM feeds, on which she continued throughout her admission and discharge home. She was extubated onto continuous positive airway pressure (CPAP) support at 18 days of life, 4 days after starting treatment. Her cerebrospinal fluid (CSF) parameters and culture was negative. Blood cultures were negative preceding and following cessation of amoxicillin.
Subsequent to this, she had four further episodes of GBS infection. Her second infection occurred at 1 month of age, required reintubation due to apnoeas and was treated with a further 14 days of antibiotics: 2 days of piperacillin/tazobactam followed by 12 days of high-dose amoxicillin (four times per day intravenous 100 mg/kg) once her blood culture result was known. On this occasion, CSF showed evidence of possible meningitis, with one white cell per mm3, 3680 red cells/mm3, a negative bacterial culture but positive PCR for GBS DNA. She received a course of dexamethasone (oral one time per day 50 microg/kg) to facilitate extubation at 6 weeks of age which was weaned and stopped over 8 weeks.
Her third infection at 8 weeks of age presented with further apnoeas and increased oxygen requirement necessitating reintubation. Her CSF was also blood-stained and GBS PCR positive, but culture negative. She was treated with a further 4 weeks of high dose amoxicillin after initial doses of empirical amoxicillin, flucloxacillin and gentamicin pending blood culture results. She was successfully extubated onto CPAP 6 days into treatment. Blood cultures were sterile 1 week into treatment and following treatment. Her fourth GBS infection occurred at 13 weeks of age, when blood cultures also isolated S. epidermidis. She was treated with amoxicillin, flucloxacillin and gentamicin for 2 days then 4 weeks of high dose amoxicillin. Patient and mother subsequently received 7 days of rifampicin (two times per day oral 10 mg/kg) simultaneously to eradicate GBS colonisation, followed by antimicrobial prophylaxis with amoxicillin for the patient. Breastfeeding was discontinued at 7 months of age.
At 17 months of age, antimicrobial prophylaxis was stopped: the patient presented 10 days following the final dose with fever, and GBS was isolated in a blood culture for the fifth time. There were no clinical features of meningitis, but lumbar puncture was not possible during this presentation, and the patient received 14 days of ceftriaxone (one time per day intravenous 100 mg/kg) followed by amoxicillin prophylaxis.
The patient’s treatment course and antibiotic use are summarised in table 1.
Summary of GBS episodes, microbiology results and treatment
GBS episode | Age | Source | Serotype | Antibiotic | Duration | Total treatment length | Comments |
1 | 14 days | Blood | Not known | Ceftazidime, vancomycin | 2 days | 14 days | |
Amoxicillin | 5 days | ||||||
Ceftazidime, vancomycin, metronidazole | 7 days | Change due to clinical suspicion of necrotising enterocolitis | |||||
2 | 4 weeks | Blood CSF (culture negative, PCR positive) |
III | Piperacillin/tazobactam | 2 days | 14 days | |
Amoxicillin | 12 days | ||||||
3 | 8 weeks | Blood CSF (culture negative, PCR positive) |
III | Amoxicillin, flucloxacillin, gentamicin | 2 days | 28 days | |
Amoxicillin | 26 days | ||||||
4 | 13 weeks | Blood | III | Amoxicillin, flucloxacillin, gentamicin | 2 days | 28 days | Staphylococcus epidermidis also isolated in blood Treatment followed by maternal and infant eradication therapy with rifampicin, and subsequent amoxicillin prophylaxis |
Amoxicillin | 26 days | ||||||
5 | 17 months | Blood | III | Ceftriaxone | 14 days | 14 days | Preceded by cessation of antimicrobial prophylaxis |
-
CSF, cerebrospinal fluid; GBS, group B streptococcus.
Investigations
Urine, and stool cultures were negative in all episodes, and there was no clinical suspicion of bone or joint infection. Blood culture isolates from episodes 2, 3, 4 and 5 were all of serotype III, all sensitive to amoxicillin. Whole body MRI (including STIR sequence), cranial ultrasound and abdominal ultrasound following her fourth episode of GBS infection showed no infective foci and a normal spleen.
Full blood count, blood film and lymphocyte subsets, including naïve T-lymphocytes, at 36 weeks’ gestation were normal. Maternal EBM was culture negative for GBS but positive on 16S PCR prior to treatment. She was discharged home at 40+2 weeks, breastfeeding on demand, on 0.2 L per min of supplemental oxygen and amoxicillin (two times per day oral 125 mg). She received primary immunisations without complications, and with satisfactory responses to Haemophilus influenzae type B and tetanus toxoid. Immunoglobulin concentrations were low (IgA<0.05 g/L, IgG 1.7 g/L and IgM 0.38 g/L) at 7 months of age (3 months corrected for gestation) but normalised aside from IgA at 2.5 years of age (IgA 0.12 g/L, IgG 5.2 g/L and IgM 0.58 g/L) consistent with exaggerated transient hypogammaglobulinaemia of infancy. Her IgG response to the 13-valent pneumococcal conjugate vaccine was poor (1/9 protective to conjugate vaccine serotypes, followed by 8/9 responsive after booster vaccination). Normal complement activation and shedding of CD62 ligand to stimulation with lipopolysaccharide and CL097 were seen, implying appropriate signalling through toll-like receptors −4 to –7 and −8. Neutrophil phagocytic function, assessed via the nitro-blue tetrazolium test and oxidation of dihydrorhodamine-123, was normal.
Outcome and follow-up
Our patient is 2.5 years old, and remains well on antimicrobial prophylaxis with amoxicillin. Despite her extreme prematurity and turbulent early life course, she is developing appropriately for age and without any other significant infections.
Discussion
Invasive GBS infection remains a significant cause of neonatal morbidity and mortality in the UK, and this is amplified in both recurrent disease and prematurity. Prematurity, low birth weight, maternal colonisation and multiple-gestation pregnancies are all risk factors for invasive GBS.8 9 GBS meningitis is associated with significant neurodevelopmental impairment10 and preterm infants are both at greater risk of developing invasive GBS and also at greater risk of consequent neurodevelopmental impairment.11
Our patient presents with five discrete episodes of GBS infection, the most reported to date and aligns with previously described characteristics including serotype, prematurity and timing of first and second episodes (mean: 10 days and 42 days of life, respectively; our patient: 14 days and 38 days of life4). Our patient was infected with GBS serotype III and GBS was isolated from breastmilk, although our patient’s fifth episode recurred 1 year after discontinuation of breastfeeding
The role of breastmilk in recurrent GBS is controversial. Maternal–infant pairs often isolate identical GBS serotypes, while GBS presence in breastmilk is associated with both mastitis and higher levels of GBS colonisation in infants. The mechanisms for GBS transmission via breastmilk are unclear and different mechanisms have been proposed, including retrograde milk flow in mammary ducts, colonised neonatal oral cavities and entero-mammary circulation.4 12 While plausible, current evidence consists of case reports and small series. A recent retrospective study of 98 cases in Italy investigated the role of maternal GBS carriage in LOD and showed that a significant proportion of LOD is associated with maternal GBS carriage, positive breastmilk culture and GBS bacteriuria.13 Our patient’s mother was not screened for GBS carriage in the antenatal period, in line with UK guidance. Estimated rates of GBS carriage in breastmilk (0.4%–3.8%) are higher than the incidence of recurrent LOD, suggesting that the presence of GBS in breastmilk is insufficient to cause disease.4 There is no consensus on screening breastmilk for GBS in recurrent disease, due to the limitations of our understanding of its role and the established benefits of breastmilk in the sick neonate. In our patient, breastmilk was only positive after the fourth episode, and only through identification of GBS DNA by 16S PCR. Decolonisation of infant mucosae and breastmilk can be achieved with concurrent maternal–infant rifampicin,4 which was well tolerated in our patient and mother, though clearly not sufficient to prevent recurrence of disease in this case.
Breastmilk as a source of recurrent infection in this infant seems unlikely as the fifth episode of GBS infection occurred after breastfeeding had stopped for more than a year. This is consistent with other case reports of recurrent GBS infection despite discontinuation of breastfeeding.14 Persistent infection is also unlikely in this case; though one study identified short courses of antibiotics are associated with GBS recurrence,5 our patient received appropriate length of treatment (≥14 days, covering for possibility of meningitis) for each episode demonstrating clinical improvement, normalisation of laboratory indices and sterile blood cultures between episodes. Imaging to identify a persistent focus for haematogeneous spread, such as an intracranial collection or soft tissue abscess, was also negative.
Prematurity increases the risk of invasive disease, case fatality and recurrence1 8 and increases the risk of neurodevelopmental impairment associated with GBS infection.11 In addition to the iatrogenic risks of prolonged hospital admission, indwelling devices and gut dysbiosis from antibiotic use, the neonatal immune system is immature, causing susceptibility to infection and impacting both innate and adaptive arms of immunity.15 Neonates depend on transplacental transfer of maternal IgG for passive immunity; this occurs largely in the third trimester, leaving preterm infants relatively hypogammaglobulinaemic.15 Furthermore, IgG repertoire diversity is reduced in preterm infants compared with term infants,15 resulting in attenuated responses to vaccine-strain antigens, as seen in our patient. Use of corticosteroids, such as the course of dexamethasone our patient received between episodes 2 and 3, might also impact the response to infection. Aberrant adaptation of GBS to its infant host might also occur by antibiotic use: treatment of mothers perinatally alters the infant microbiome by reducing Bifidobacteria species,5 while treatment of infants may cause emergence of hypervirulent strains such as CC176 and impact on development of host immunity. Importantly, extreme prematurity per se as well as the prolonged course of broad-spectrum antibiotics given will have significantly increased the risk of infection and sepsis in this infant.16
Compared with other encapsulated bacteria, there are few associations between GBS and inborn errors of immunity. Case reports describe recurrent or disseminated GBS in patients with variants in C2,17 IRAK4 and MyD88;18 these were all excluded in our patient given intact complement activation and appropriate CD62L shedding, along with analysis of the NHS Genomic Service R15 panel targeted to inborn errors of immunity in the patient and her parents finding no pathogenic variants. In the absence of a clear driver for our patient’s recurrent GBS disease and recrudescence when prophylaxis was discontinued, she remains on amoxicillin. Despite her turbulent early life course and prematurity, she is well, free of other infections, and her development is normal.
Learning points
-
Group B streptococcus (GBS) infection is an important cause for neonatal disease and can be early-onset (<7 days of life) or late-onset (7–89 days of life).
-
GBS infection is associated with significant neurodevelopmental impairment, a risk which is increased by prematurity (<37 week gestation).
-
Recurrent GBS occurs in 0.5%–3% of late-onset cases and prematurity is a risk factor for recurrent GBS due to premature infants typically having prolonged hospital admissions with indwelling devices but also having an immature innate and adaptive immune system.
-
The role of breastmilk in recurrent GBS is controversial but mother infant pairs do often isolate identical GBS serotypes as described in this case report.
-
There is little systematic evidence for the management of recurrent GBS although concurrent treatment of infant mucosae and breastmilk with well-tolerated agents such as rifampicin is well described.
Ethics statements
Patient consent for publication
References
Footnotes
-
ISvdL and CT are joint first authors.
-
Twitter @christotsilifis
-
Contributors The following authors were responsible for drafting of the text, sourcing and editing of clinical images, investigation results, drawing original diagrams and algorithms and critical revision for important intellectual content: ISvdL, CT, KA and EW. ISvdL and CT contributed equally to this manuscript and are joint first authors. The following authors gave final approval of the manuscript: ISvdL, CT, KA and EW.
-
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
-
Case reports provide a valuable learning resource for the scientific community and can indicate areas of interest for future research. They should not be used in isolation to guide treatment choices or public health policy.
-
Competing interests None declared.
-
Provenance and peer review Not commissioned; externally peer reviewed.
- © BMJ Publishing Group Limited 2023. No commercial re-use. See rights and permissions. Published by BMJ.
Use of this content is subject to our disclaimer