Neonatal Group B Streptococcus Disease

Abstract

Group B Streptococcus (GBS) is an important cause of neonatal sepsis among both term and preterm infants. Because GBS colonizes human genitourinary and gastrointestinal tracts, a significant focus of neonatal GBS disease prevention is to interrupt vertical transmission of GBS from mother to infant during parturition. Routine antepartum GBS screening in pregnant women, as well as widespread use of intrapartum antibiotic prophylaxis, have aided in overall reductions in neonatal GBS disease over the last three decades. However, neonatal GBS disease persists and may cause mortality and significant short- and long-term morbidity among survivors. Here, we highlight contemporary epidemiology, microbial pathogenesis, and the clinical presentation spectrum associated with neonatal GBS disease. We summarize obstetric recommendations for antenatal GBS screening, indications for intrapartum antibiotic prophylaxis, and considerations for antibiotic selection. Finally, we review national guidelines for risk-assessment and management of infants at risk for GBS disease.

INTRODUCTION

Group B Streptococcus (GBS) emerged in the 1970s as a leading cause of neonatal sepsis in the United States, and currently accounts for approximately 30% of early-onset neonatal sepsis cases¹. Maternal GBS colonization occurs in 10–35% of parturients in the U.S.^2–4, and is a major risk factor for neonatal early-onset GBS disease acquired via vertical transmission during birth. The overall incidence of neonatal GBS disease has decreased over the last several decades, largely due to reductions in early-onset GBS disease in the era of protocolized screening for maternal GBS colonization and widespread use of GBS-targeted intrapartum antibiotic prophylaxis (IAP). However, there are no current strategies for late-onset neonatal GBS disease prevention, and the national incidence of late-onset infection has not changed with use of IAP. This article summarizes the epidemiology of neonatal GBS disease, consensus recommendations for maternal screening and intrapartum antibiotic prophylaxis, and management of neonates at risk for GBS disease.

EPIDEMIOLOGY

Early-Onset GBS Disease

Early-onset GBS disease (EOGBS) is defined as infection presenting between days 0–6 after birth⁵. Estimated U.S. EOGBS incidence has steadily decreased over the past three decades^6–8: from 1.8 cases/1000 live births in 1992⁹ to 0.2 cases/1000 live births in 2020¹⁰. Globally, EOGBS incidence is 0.41 cases/1000 live births, with 2-fold higher incidence in low- and middle-income compared to high-income countries¹¹. In 2020, 43% of all neonatal GBS disease cases in the U.S. occurred in the early-onset period¹⁰. A Neonatal Research Network (NRN) study of infants born 2015–2017 with early-onset neonatal sepsis (EOS) identified that GBS caused 30% of all EOS cases occurring within 72 hours after birth. GBS was the second-most common pathogen overall in this cohort (after Escherichia coli, which was isolated in 35% of all cases)¹. However, early-onset GBS disease rates differ between infants born at term versus preterm; almost 75% of all EOGBS cases occur in term infants⁶. Among infants within the NRN with EOS, GBS accounted for 51% of EOS cases in term infants (born ≥37 weeks’ gestation), compared to 13% of EOS cases in preterm infants (born ≤36 weeks gestation). Among very low birth weight infants (born <1500 grams) admitted to Vermont Oxford Network center NICUs in 2018–2019, GBS accounted for 19% of EOS cases¹².

Late-Onset GBS Disease

Late-onset GBS disease (LOGBS) is defined as invasive GBS disease presenting on days 7–89 after birth. LOGBS incidence has remained stable over the past two decades, with incidence estimated at 0.3–0.4 cases/1000 live births both in the U.S. and worldwide^6,8,11. LOGBS is now the predominant form of neonatal GBS disease, accounting for 57% of all neonatal GBS cases in 2020¹⁰. Preterm infants (born <37 weeks’ gestation) account for 10–11% of all US births¹³ but Centers for Disease Control and Prevention (CDC) surveillance reports that 48% of LOGBS cases occur among preterm infants⁶. Population-level estimates suggest that preterm infants are at six times higher risk of developing LOGBS, compared to term infants⁶. Accordingly, a substantial proportion of LOGBS occurs in chronically hospitalized infants, accounting for 19% of all LOGBS cases in a large Japanese surveillance study¹⁴.

PATHOGENESIS

Microbiology

GBS is a beta-hemolytic encapsulated Gram-positive bacterium that colonizes the human genitourinary and gastrointestinal tracts. Colonization is mediated by GBS surface proteins which allow effective adherence to the extracellular matrix on the epithelial surface¹⁵. Molecular pathogenesis of GBS is complex¹⁵ and involves invasion of epithelial barriers (including placental membranes and pulmonary epithelium) via endocytosis. GBS produces numerous toxins (e.g., beta-haemolysin/cytolysin and CAMP factor) that proceed to cause direct cytotoxic tissue injury. GBS evades the host immune response via its polysaccharide surface capsule, which mimics human surface antigens and evades complement deposition, thus interfering with host opsonization and phagocytosis. Among bacteria undergoing phagocytosis, GBS is uniquely able to resist oxidative burst killing via expression of superoxide dismutase and production of carotenoid pigments that neutralize host oxidative stressors.

GBS’ capsular polysaccharides are major virulence factors and the primary microbial element distinguishing GBS serotypes¹⁶. Ten GBS serotypes have been identified (types Ia, Ib, and II-IX), with types Ia, Ib, II, III, and V occurring most commonly and accounting for 98% of serotypes causing maternal colonization and neonatal GBS disease^6,11,16,17. Serotype distribution differs by timing of GBS onset; CDC US surveillance during 2006–2015 identified that EOGBS was most commonly associated with serotypes Ia (27%), III (27%), and II (16%), while LOGBS was most commonly associated with serotypes III (56%) and Ia (20%)⁶.

GBS is a common cause of neonatal meningitis and effectively invades and causes direct injury to the central nervous system. GBS penetrates the epithelial blood-brain barrier via transcytosis, aided by beta-haemolysin/cytolysin’s direct cytotoxic activity. Preclinical studies in mice suggest that invasion of GBS through the blood-brain barrier then triggers a robust inflammatory response leading to diffuse neutrophil recruitment, cytokine release, vascular damage, and neuronal death¹⁵.

Transmission

Vaginal-rectal GBS colonization rates range from 10–35% among parturients in the U.S.^2–4,18. Maternal GBS colonization can produce an adaptive immune response in the form of serotype-specific IgG production, detectable in 50% of colonized mothers at delivery². Among colonized women, young maternal age is associated with negative GBS serostatus. Neonates colonized at birth almost universally share the same GBS serotype colonizing their mothers antenatally^19–21. Among colonized mothers not receiving intrapartum antibiotic prophylaxis (IAP), neonatal GBS colonization rates at birth are reported at 43–60%^22,23, decreasing to as low as 0–9% in the setting of intrapartum ampicillin treatment^23–25. Maternal receipt of IAP is also associated with lower rates of heavy neonatal GBS colonization burden (decreasing from 35% to 12% in one prospective cohort study of mother-infant dyads)²¹.

Among all infants of GBS-colonized women, EOGBS rates are approximately 1–2% in the absence of IAP, decreasing to 0.3% in settings with routine IAP use²⁶. Neonatal EOGBS results from vertical transmission of GBS from mother to fetus via ascending infection through ruptured membranes, translocation through intact membranes, or aspiration during parturition²⁵. Higher maternal colonization density is associated with increased likelihood of neonatal colonization at birth^24,27. In addition to GBS colonization, risk factors for development of EOGBS include preterm birth, prolonged duration of membrane rupture, presence of intra-amniotic infection, GBS bacteriuria during pregnancy, and history of a prior neonate with GBS disease²⁸. Transplacental transfer of maternal serotype-specific GBS antibodies appears to protect neonates from invasive GBS disease. Presence of serotype-specific maternal antibody is associated with an estimated 70–90% risk reduction of neonatal EOGBS²⁹, while negative maternal GBS serostatus is associated with invasive neonatal GBS disease³⁰. In the era of universal screening and GBS IAP, EOGBS now most commonly occurs in the setting of negative maternal GBS screening results (reported 50–80%)^1,31,32 – potentially attributable to false-negative antenatal screening results, or to maternal GBS colonization occurring after screening but prior to delivery.

GBS acquisition in LOGBS is likely multifactorial. Only 10–50% of infants with LOGBS are born in the setting of positive antenatal maternal GBS screening^33–35. Nonetheless, vertical transmission is implicated in LOGBS, with up to 50% of infants colonized at birth¹⁸. Delayed vertical transmission from GBS-colonized mothers to their infants after hospital discharge also occurs²¹. IAP reduces maternal GBS colonization burden and decreases immediate transmission during birth; delayed transmission can occur when maternal GBS colonization burden rises after IAP-induced suppression wanes. Mothers with negative antenatal GBS screens may become colonized after delivery, resulting in delayed neonatal GBS colonization and disease onset. Among 53 Italian neonates with LOGBS, only 33% were born in the setting of positive maternal GBS screening; by the time LOGBS was diagnosed, 63% of these infants’ mothers were GBS-colonized³³. A longitudinal cohort study of infants of GBS-colonized mothers identified that (1) delayed neonatal colonization does occur, as early as 15 days post-delivery, and (2) GBS strains identified in delayed neonatal colonization were identical to maternal strains present at birth²¹. Associations of LOGBS with exposure to GBS from breastmilk are reported but the importance of breastmilk as a source of neonatal colonization is controversial. One prospective study of LOGBS found 6/83 mothers GBS-colonized postpartum had GBS in breastmilk²¹. It remains unclear if the colonized or infected mammary gland is a source of neonatal colonization, or vice versa^33,36–38. Finally, horizontal transmission is implicated, both from community nonmaternal sources and from nosocomial origins (including hospital-based outbreaks with presumed transmission via nursery staff)^33,39.

CLINICAL PRESENTATION

GBS is primarily a perinatal pathogen, and in addition to neonatal disease it causes bacteremia, chorioamnionitis, and endometritis in parturient women⁴⁰. Rising rates of invasive GBS disease are also noted in non-pregnant adults, particularly the elderly, largely manifested as bacteremia, pneumonia, and skin/soft tissue infections^41,42.

Fetal

GBS infection is associated with adverse fetal outcomes, particularly with stillbirth. A systematic review identified that 0–12% of stillbirths are associated with isolation of GBS from an otherwise sterile compartment⁴³ - an estimated annual 57,000 GBS-associated stillbirths worldwide⁴⁴. Maternal GBS colonization has also been associated with preterm birth, though this relationship is confounded by accompanying perinatal factors (e.g., preterm labor, preterm premature rupture of membranes) that may themselves increase the risk of GBS infection. Maternal GBS bacteriuria is associated with both preterm labor and premature rupture of membranes, but an association with preterm birth has not been consistently identifed^45,46

Early-onset GBS Disease

While EOGBS is defined as GBS disease presenting during days 0–6 after birth, the vast majority of contemporary EOGBS cases (95%) present during the birth hospitalization, within the first 48 hours after birth⁶. Clinical risk factors for EOGBS overlap with general neonatal early-onset sepsis risk factors, and include preterm birth, maternal intrapartum fever, clinical concern for intraamniotic infection, prolonged duration of membrane rupture, and inadequate exposure to IAP (if indicated)^31,47. Isolated bacteremia is the most common manifestation of EOGBS (80–90%); meningitis with cerebrospinal fluid (CSF) culture-confirmed infection is less commonly reported (1–10%)^1,6,7. Early-onset GBS meningitis incidence may be underestimated given low rates of lumbar punctures performed prior to antibiotic initiation¹. Clinical signs at EOGBS presentation are nonspecific and commonly include respiratory distress, tachycardia, apnea, and/or lethargy. Infants with early-onset meningitis may present with seizures⁴⁸. Rarer but severe manifestations of EOGBS disease may include shock, perinatal encephalopathy, and pneumonia with or without pulmonary hypertension⁴⁹. GBS pneumonia may appear radiographically indistinguishable from the respiratory distress syndrome of the newborn^50,51, mandating a high index of suspicion for early-onset infection among infants with respiratory failure at birth.

Late-onset GBS Disease

Reported median age at LOGBS presentation ranges from 27–41 days old, with term infants presenting earlier than preterm infants^6,14,33,35. Clinical factors associated with LOGBS include preterm birth, maternal GBS colonization, and multiple gestation^33,52. Globally, human immunodeficiency virus (HIV)-exposed but uninfected infants appear to have four-fold higher rates of LOGBS compared to HIV-unexposed infants; reduced transplacental transfer of GBS-specific antibodies to HIV-exposed infants is one proposed explanation for this finding⁵³. Clinical manifestations of LOGBS are more heterogeneous compared to EOGBS. Isolated bacteremia occurs most frequently (61%), though meningitis is also common (31%)⁶. Less common presentations include pneumonia, urinary tract infection, septic arthritis/osteomyelitis, skin and soft tissue infections, and adenitis^6,33,35. Importantly, up to 8% of infants with any LOGBS disease (and up to 30% of infants with CSF culture-confirmed meningitis) have negative blood cultures^6,54, highlighting the utility of inclusion of urine, CSF, and focal site cultures (as applicable) in routine evaluations of infants with suspected late-onset neonatal sepsis. Clinical signs reported in LOGBS include fever, poor feeding, respiratory distress with tachypnea, emesis, changes in mental status (irritability, lethargy), and seizures (particularly among infants with LOGBS meningitis)^35,48,49. LOGBS may progress to fulminant disease with septic shock (variably reported with rates ranging 3–10%)^6,33,35, and up to 30% of all infants with LOGBS presenting to emergency departments require intensive care unit admission³⁵. Table 1 summarizes characteristics of EOGBS and LOGBS.

Table 1:

Comparison of Early- and Late-Onset GBS Disease

	Early-Onset GBS	Late-Onset GBS
Timing	0 – 6 days of age	7 – 89 days of age
Epidemiology	U.S. incidence decreasing, currently ~0.2 cases/1000 live births ~75% of early-onset GBS occurs in term infants (≥37 weeks gestation) Prevalence in preterm infants with early-onset sepsis is 13–19%	Stable U.S. incidence (~0.3 cases/1000 live births) ~50% of late-onset GBS occurs in term infants Most common form of neonatal GBS disease (57% of all neonatal GBS cases)
Transmission	Vertical transmission from a colonized mother	Multifactorial: vertical transmission; horizontal transmission via nonmaternal community and/or nosocomial sources
Risk factors	Maternal GBS colonization Preterm labor Prolonged membrane rupture Concern for intra-amniotic infection at delivery	Maternal GBS colonization Prematurity
Clinical Presentation	Usually presents within first 48 hours after birth Isolated bacteremia in 80–90%, meningitis in up to 10% Nonspecific presenting signs, less frequent severe presentations (pneumonia, pulmonary hypertension, shock)	Median ages at presentation 27–41 days of age Isolated bacteremia in 60%, meningitis in 30–35%. Less commonly presents as pneumonia, UTI, bone/joint infection, skin/soft tissue infection Nonspecific presenting signs; occasionally proceeds to shock
Management	Perform lumbar puncture if not already done Antibiotic of choice is parenteral penicillin G (ampicillin is an alternative agent) Antibiotic treatment duration 10 days for isolated bacteremia; 14 days for meningitis. Longer courses required for osteomyelitis, ventriculitis. Hearing screening in all infants with meningitis; consider cranial imaging also

Recurrent GBS disease is uncommon and poorly described, though appears to occur among 2–3% of infants with any invasive GBS disease. This phenomenon most frequently manifests as LOGBS in both the initial and the recurrent episodes, and most commonly affects preterm infants^14,33,55. Recurrent GBS disease may occur in the setting of recrudescence of the original GBS serotype in a colonized infant; most studies of recurrent disease find identical GBS strains in each episode of disease^55,56. In one study of 21 infected infants and their 20 mothers, oropharyngeal and gastrointestinal GBS colonization was present in half of the infants at the completion of appropriate intravenous antibiotic therapy. Rifampin was administered to infants with persistent colonization and to their mothers, but this treatment failed to eradicate colonization in the majority of the dyads⁵⁷. The association of GBS-positive breast milk with recurrent GBS disease is controversial. In a case series of 48 LOGBS cases associated with GBS-positive breast milk, 35% of cases were recurrent³⁸. However, breast milk is also an important source of protection from neonatal GBS disease (via maternal serotype-specific GBS antibodies, if present), and it is unclear whether GBS-positive breast milk causes LOGBS or if it merely reflects a site of maternal colonization⁴⁹. Horizontal transmission may account for episodes of recurrent disease with discordant GBS serotypes^14,55. Immature adaptive immune responses reducing neonatal GBS antibody production may also compound susceptibility to recurrent infection.

Very Late-Onset GBS Disease

“Very late-onset” GBS disease in infancy presents on or after day 90 of age. Very-late onset GBS disease accounts for 10–25% of all GBS cases presenting after 7 days of age^34,54,58. Median ages at presentation are 115 days (range 91–226 days)⁵⁴ and 118 days (interquartile range 98–790 days)³⁴. Similar to LOGBS, bacteremia and meningitis are the primary clinical disease manifestations. Preterm birth is a consistent risk factor^{14,54,58–60}. A national surveillance study of 242 late-onset GBS meningitis cases in France identified a significantly higher prevalence of infants born <32 weeks gestation among infants with very late-onset disease (32%) compared to late-onset disease (7%)⁵⁴. The vulnerability of very preterm infants may be related to reduced exposure to protective, transplacentally-acquired GBS-specific antibodies as well as to prolonged immature innate immune responses. Innate or acquired immunodeficiency is also reported in the setting of very late-onset GBS disease, including hypogammaglobulinemia, symptomatic HIV infection, and immunosuppression following organ transplant^14,34,59.

OBSTETRIC SCREENING AND INTRAPARTUM MANAGEMENT

In 2020, the American College of Obstetrics and Gynecology (ACOG) published updated consensus guidance for obstetric approaches to neonatal GBS disease prevention⁶¹. This guidance is detailed below, and summarized in Table 2.

Table 2:

ACOG Guidance for Group B Streptococcus Screening and Intrapartum Antibiotic Prophylaxis

Screening
Eligibility and timing	All parturients between 36 0/7 and 37 6/7 weeks’ gestation
Method	Vaginal-rectal cultures are gold standard Nucleic acid amplification testing (NAAT) may be used as an alternative, but cannot report antibiotic susceptibilities
Intrapartum Antibiotic Prophylaxis (IAP)
Indications for IAP	Positive GBS culture on or after 36 0/7 weeks’ gestation (unless delivery via Cesarean section without labor or rupture of membranes) Positive GBS culture obtained <36 0/7 weeks’ gestation, if delivery occurs within 5 weeks of the date of screening GBS bacteriuria during the current pregnancy Prior neonate with invasive GBS disease Unknown GBS status and preterm prelabor rupture of membranes at <37 0/7 weeks Unknown GBS status at labor onset, and at least one of: Gestation <37 0/7 weeks (preterm labor) Duration of membrane rupture >18 hours Intrapartum temperature >100.4 °F GBS positive in a prior pregnancy Intrapartum NAAT positive Preterm premature rupture of membranes
Antibiotic Selection
No history of penicillin allergy	Penicillin G (intravenous, IV) is agent of choice, start at labor onset and continue through delivery Ampicillin IV is acceptable alternative
History of penicillin allergy	If history of anaphylaxis/IgE mediated hypersensitivity reaction, send clindamycin susceptibility testing on GBS isolate If clindamycin-susceptible: clindamycin IV recommended If clindamycin resistant: vancomycin IV recommended If allergy history not consistent with IgE-mediated hypersensitivity reaction, cefazolin IV recommended

Antepartum Screening

ACOG recommends universal GBS screening among all pregnant women between 36 0/7 and 37 6/7 weeks’ gestation, maximizing the likelihood that GBS screening results will reflect colonization status at delivery. Recto-vaginal cultures are the current gold-standard screening test, although nucleic acid amplification testing (NAAT) demonstrates similar-to-improved rates of GBS isolation. Limitations of NAAT for routine screening include an inability to perform antibiotic susceptibility testing on positive GBS screens, and up to 10% false-negative rate when used as a rapid point-of-care test.

Indications for Intrapartum Antibiotic Prophylaxis

IAP is administered with the intent of suppressing colonizing GBS bacterial burden at delivery to reduce neonatal colonization and EOGBS disease. IAP is indicated for all parturients with positive GBS screening results obtained after 36 0/7 weeks gestation. Additional indications for IAP, including in the setting of unknown GBS status, are listed in Table 2. According to contemporary CDC surveillance, approximately 50% of EOGBS cases occur in settings where perinatal IAP was not indicated, largely due to negative antenatal GBS screening results⁶. In 20% of EOGBS cases, IAP was indicated but not administered. Less than 10% of cases occurred in the setting of receipt of IAP for >4 hours prior to delivery as recommended by ACOG guidance, and in these cases it was unclear if the administration of alternate medications such as clindamycin were given appropriately.

Antibiotic Selection and Timing of Administration

Intravenous penicillin is the preferred agent for GBS-targeted IAP given its narrow spectrum of activity with exceedingly low GBS resistance rates; ampicillin is an acceptable alternative (Table 2). These agents should be dosed every 4 hours until delivery; the effectiveness of penicillin or ampicillin to prevent neonatal EOGBS is highest when prophylaxis is received at least 4 hours prior to delivery⁶². IAP exposure as soon as 1–2 hours prior to delivery does reduce neonatal GBS colonization rates, though impact on EOGBS reduction is more difficult to ascertain⁶³.

Approximately 10% of U.S. parturients report a penicillin allergy³²; clindamycin susceptibility testing should be performed on GBS isolates from penicillin-allergic parturients. In the setting of “high-risk” maternal penicillin allergy (e.g., presumed IgE-mediated hypersensitivity reaction), IAP should consist of clindamycin if the GBS isolate is clindamycin-sensitive. If the isolate is clindamycin-resistant, IAP should consist of vancomycin. Among women with “low-risk” penicillin allergy (history not consistent with acute hypersensitivity reaction), ACOG recommends cefazolin as GBS IAP.

Fetal-Neonatal Exposure to Intrapartum Antibiotics

Transplacental fetal-neonatal exposure to maternally-administered antibiotics depends on multiple factors, including maternal drug dosing, duration of exposure prior to delivery, and conditions affecting placental blood flow and efficiency of drug transfer (e.g., pre-eclampsia, diabetes)⁶⁴. Beta-lactam antibiotics (penicillin, ampicillin, and cefazolin) exhibit rapid placental transfer; GBS bactericidal concentrations are achieved in the fetal compartment within several minutes of administration and can persist for several hours^64,65. While available data suggest transplacental vancomycin and clindamycin transfer occurs, the magnitude of fetal exposure and its effectiveness against EOGBS is unclear. Vancomycin undergoes renal elimination (which is physiologically accelerated during pregnancy) and transplacental transfer is slow owing to its large molecular weight⁶⁴. Ex vivo placental models suggested limited placental transfer of vancomycin at standard dosing (1 gram every 12 hours)^66,67. Weight-based vancomycin dosing (20 mg/kg every 8h) more frequently produced therapeutic vancomycin levels in umbilical cord blood compared to standard dosing (77–83% vs 9%, respectively)^68,69. Based on these data, obstetric GBS IAP guidelines were revised to recommend weight-based vancomycin dosing in 2019⁶¹. Clindamycin is highly protein-bound, and transplacental transfer is three times slower than beta-lactam antibiotics⁶⁴. Studies of transplacental transfer of clindamycin provide conflicting results; one pharmacokinetic study of 7 dyads identified that maternal clindamycin concentrations may be subtherapeutic, with neonatal clindamycin exposure even lower⁷⁰. However, another study among 23 term infants identified that 96% achieved therapeutic clindamycin concentrations in cord blood following intrapartum exposure⁷¹. Further pharmacokinetic data are needed to optimize IAP antibiotic dosing, particularly among preterm infants for whom this data is most lacking.

GBS antibiotic resistance patterns have evolved since its emergence in the 1970s. Prevalence of erythromycin resistance among maternal GBS isolates is 33–45%^6,72,73, and erythromycin is no longer recommended as GBS IAP as of 2010⁷⁴. Among currently-utilized alternative IAP agents, clindamycin resistance rates in maternal GBS isolates are 18–33%, while vancomycin resistance is rare^6,72,73. Rising resistance is similarly seen in invasive neonatal GBS disease, with clindamycin resistance in 50% and erythromycin resistance in 60% of infecting neonatal GBS isolates in 2021⁷⁵. GBS remains highly susceptible to first-line IAP agents, and resistance to penicillin or ampicillin is rare.

MANAGEMENT OF NEONATES AT RISK FOR EARLY-ONSET GBS DISEASE

Risk Assessment

All newborn infants require a clinical assessment of their early-onset sepsis risk at birth. The American Academy of Pediatrics (AAP) endorses multiple approaches to assess early-onset sepsis risk (EOS; inclusive of EOGBS risk) among infants born ≥35 weeks gestation (Figure 1)^49,76. The first approach is a categorical risk assessment, which identifies infants at higher EOS risk based on clinical status, presence of maternal intrapartum fever, and exposure to appropriate GBS IAP (if indicated). The Neonatal Early-Onset Sepsis Calculator (https://neonatalsepsiscalculator.kaiserpermanente.org) is a web-based multivariate approach to risk assessment that calculates a posterior probability of EOS based on gestational age, highest maternal intrapartum temperature, duration of membrane rupture, maternal GBS colonization status, maternal receipt of IAP (if indicated) and duration of fetal exposure prior to delivery, and infant clinical status. A final approach leverages serial clinical observations to identify infants who may be at higher risk of EOS. An individual center’s choice of EOS risk assessment approach may be influenced by numerous factors, including: local EOS incidence, preferences balancing risk acceptance with empiric antibiotic utilization, and availability and training of personnel to perform vital signs and serial clinical assessments in the nursery.

Comparison of American Academy of Pediatrics-endorsed approaches for early-onset sepsis risk assessment in term infants.

For the purposes of neonatal risk-stratification, AAP guidance defines “adequate” IAP as maternal administration of penicillin, ampicillin, or cefazolin, given these agents’ favorable pharmacokinetic parameters, efficiency of transplacental transfer, and low GBS resistance rates⁴⁹. Maternal receipt of clindamycin or vancomycin is considered inadequate IAP, given lack of clarity surrounding transplacental exposures and subsequent efficacy in reducing neonatal EOGBS disease.

These risk-stratification approaches do not apply to preterm infants born ≤34 weeks, who are at statistically highest risk of EOS and for EOS-related mortality. The maternal GBS status is often unknown in these infants, as antenatal GBS screening is not routinely performed until 36 0/7 weeks’ gestation. Accordingly, most (80–90%) very preterm infants in the U.S. receive empiric antibiotics at birth due to EOS risk⁷⁷. The AAP currently endorses identification of a narrow cohort of preterm infants at very low EOS risk based on delivery characteristics, and for whom empiric antibiotic therapy may be deferred in the appropriate clinical scenario⁷⁸.

Management of Infants with Suspected or Confirmed Early-Onset GBS Disease

Risk-assessment strategies are intended to support management of infants with varying levels of EOS risk at birth. Options for clinical management among infants with higher EOS risk range from enhanced clinical observation, to administration of empiric antibiotics following collection of a blood culture^49,76. Recommended empiric antibiotic regimens for infants aged 0–7 days at risk of EOGBS include ampicillin and an aminoglycoside. Among healthy infants ≥8 days of age in the, empiric antibiotic therapy should include ampicillin and a cephalosporin (ceftazidime if aged 8–28 days, ceftriaxone if aged 29–90 days). Vancomycin may be added if there is high clinical concern for meningitis. Empiric antibiotic regimens for infants at risk for invasive GBS disease who remain hospitalized beyond 72 hours of age may vary by institution and depend on local antibiograms and patient-level factors (e.g., presence of indwelling central lines).

Among infants with blood culture-confirmed GBS disease, a lumbar puncture should be performed if not already performed. Antibiotic therapy should be narrowed based on antibiotic susceptibilities; the vast majority of isolates are susceptible to parenteral penicillin G, which is the preferred agent for GBS monotherapy (ampicillin is an acceptable alternative) (Table 3). The recommended total antibiotic duration for treatment of isolated GBS bacteremia is 10 days. If meningitis is present, recommended antibiotic duration is 14 days, and clinicians should consider adjunctive brain imaging (e.g., MRI) to assess for associated abscess or ventriculitis. Longer antibiotic durations are required for GBS osteomyelitis (at least 3 weeks) and ventriculitis (at least 4 weeks); clinicians should consider consultation with a pediatric infectious disease specialist in these cases.

Table 3:

Antibiotic Regimens for Neonatal Early- and Late-Onset GBS Disease

	Antibiotic Route	Early-Onset GBS (≤7 days of age)		Late-Onset GBS (8–28 days of age)
		GA ≤34 weeks	GA ≥35 weeks	GA ≤34 weeks	GA ≥35 weeks
Bacteremia
Penicillin G aqueous	IV, IM	50,000 U/kg every 12 hours	50,000 U/kg every 12 hours	50,000 U/kg every 8 hours	50,000 U/kg every 8 hours
Ampicillin	IV, IM	50 mg/kg every 12 hours	50 mg/kg every 8 hours	75 mg/kg every 12 hours	50 mg/kg every 8 hours
Meningitis
Penicillin G aqueous	IV, IM	150,000 U/kg every 8 hours	150,000 U/kg every 8 hours	125,000 U/kg every 6 hours	125,000 U/kg every 6 hours
Ampicillin	IV, IM	100 mg/kg every 8 hours	100 mg/kg every 8 hours	75 mg/kg every 6 hours	75 mg/kg every 6 hours

Adapted from Table 4.2, Antibacterial Drugs for Neonates (≤28 Postnatal Days of Age). In: David W. Kimberlin, MD, FAAP, Elizabeth D. Barnett, MD, FAAP, Ruth Lynfield, MD, FAAP, Mark H. Sawyer, MD F, eds. Red Book: 2021–2024 Report of the Committee on Infectious Diseases. 32nd ed. Itasca, IL: American Academy of Pediatrics; 2021: 877.

GA, gestational age (weeks); IM, intramuscular; IV, intravenous; mg/kg, milligrams per kilogram; U/kg, units per kilogram

PROGNOSIS

EOGBS case-fatality rates are estimated at 5–7%^1,6,8, with preterm infants at much higher risk of death than term infants (19% vs 2% mortality rates, respectively)⁶. LOGBS case-fatality rates are estimated at 3–5%, with higher mortality rates among preterm (8%) compared to term (3%) infants^6,8,33. Infants surviving invasive GBS infection are at increased risk for neurodevelopmental impairment (NDI), which may include intellectual, motor, visual, and/or hearing impairments. In a systematic review, the prevalence of any NDI among survivors of neonatal GBS meningitis was 32% at 18 months’ follow-up, with moderate-severe NDI in 18% of survivors⁷⁹. Clinical factors associated with death or NDI in GBS meningitis include higher illness severity at presentation, failed hearing screening at hospital discharge, abnormal neurologic examination at discharge, and abnormal brain imaging after completion of antibiotic therapy⁸⁰. Extremely preterm infants, who are already at significant baseline risk of death or NDI, may be especially vulnerable to poor outcomes from invasive GBS disease. Among infants born 22–28 weeks gestation with invasive GBS disease in the Neonatal Research Network, 79% died or survived with NDI: the risk of death or NDI was 1.2-fold higher compared to extremely preterm infants with non-GBS infections, and 1.44-fold higher compared to uninfected infants⁸¹.

FUTURE DIRECTIONS

Neonatal GBS disease is not eradicated with universal GBS screening and IAP utilization, and LOGBS disease rates remain stable. This has spurred efforts to develop a maternal GBS vaccine, with aims to (1) reduce maternal GBS colonization and (2) reduce incidence of neonatal GBS disease through maternal antibody transfer via the placenta and/or breastmilk. One candidate is a type III capsular polysaccharide-tetanus toxoid conjugate vaccine; in a phase 2 randomized controlled trial, this vaccine induced significantly higher GBS type-III specific IgG and significantly delayed maternal GBS colonization with GBS type III, compared to women receiving control tetanus toxoid only⁸². Another candidate is a hexavalent GBS conjugate vaccine targeting the most frequent disease-causing GBS serotypes: Ia, Ib, II, III, IV, and V. In phase 1/2 trials, this vaccine induced robust immunoglobin responses that persisted for at least 6 months⁸³. Other candidate vaccines target protein subunits, specifically alpha-like proteins (Alp) associated with the GBS capsular surface^84,85. These candidate GBS vaccines require significant additional scrutiny, particularly with regard to reduction of maternal colonization rates and neonatal GBS disease. Randomized controlled trials may require use of surrogate clinical endpoints for neonatal disease, given that the relatively low incidence of EOGBS disease would require very large trial enrollment to demonstrate vaccine efficacy. These surrogate endpoints may include correlates of protection present in maternal and neonatal sera⁸⁶.

1. Understand the epidemiology of early- and late-onset neonatal GBS disease.

2. Understand the pathogenesis of perinatal GBS disease

3. Review obstetric guidelines for GBS screening and intrapartum antibiotic prophylaxis.

4. Understand approaches to management of infants at risk of GBS disease, including postnatal risk stratification and management of confirmed infections

EVIDENCE/SUMMARY.

• Invasive neonatal GBS disease now most frequently occurs as a late-onset process (on or after day 7 of age); incidence of early-onset GBS disease (occurring during days 0–6 of age) has declined in the setting of contemporary intrapartum antibiotic prophylaxis utilization. (Based on evidence from research)

• Early-onset GBS disease is caused by vertical transmission of GBS from colonized mothers to their infants; late-onset GBS disease may be acquired via vertical transmission or via horizontal transmission from nosocomial and/or nonmaternal community sources. (Based on strong evidence from research and expert consensus)

• Maternal receipt of GBS-specific intrapartum antibiotic prophylaxis reduces maternal GBS colonization burden and neonatal early-onset GBS disease. (Based on strong evidence from research and consensus guidelines).

• For the purposes of stratification of neonatal GBS risk at birth, exposure to intrapartum penicillin, ampicillin, or cefazolin constitutes “adequate” GBS prophylaxis. Intrapartum exposure to vancomycin or clindamycin is considered inadequate prophylaxis against neonatal GBS disease. (Based on some evidence from research and consensus guidelines).

• At birth, all newborns require an individualized clinical assessment of their risk for early-onset GBS disease. Multiple acceptable risk-assessment strategies are available to neonatal clinicians. (Based on strong evidence from research and consensus guidelines).

• Infants surviving neonatal GBS disease, particularly those with GBS meningitis, are at significant risk for neurodevelopmental sequelae and should receive close developmental follow-up. (Based on evidence from research).

Abbreviations:

AAP

American Academy of Pediatrics

ACOG

American College of Obstetrics and Gynecology

CDC

Centers for Disease Control and Prevention

EOGBS

Early-onset group B Streptococcus disease

EOS

Early-onset sepsis

GBS

Group B Streptococcus

IAP

Intrapartum antibiotic prophylaxis

LOGBS

Late-onset group B Streptococcus disease

NAAT

Nucleic acid amplification testing

NDI

Neurodevelopmental impairment

NICU

Neonatal intensive care unit

NRN

Neonatal Research Network