Etiology

Group B streptococci are facultative anaerobic gram-positive cocci that form chains or diplococci in broth and small gray-white colonies on solid medium. GBS is definitively identified by demonstration of the Lancefield group B carbohydrate antigen, such as with latex agglutination techniques widely used in clinical laboratories. Presumptive identification can be established on the basis of a narrow zone of β-hemolysis on blood agar, resistance to bacitracin and trimethoprim-sulfamethoxazole, lack of hydrolysis of bile esculin, and elaboration of CAMP factor (named for the discoverers, Christie, Atkins, and Munch-Petersen), an extracellular protein that, in the presence of the β toxin of Staphylococcus aureus, produces a zone of enhanced hemolysis on sheep’s blood agar. Individual GBS strains are serologically classified according to the presence of one of the structurally distinct capsular polysaccharides, which are important virulence factors and stimulators of antibody-associated immunity. Ten GBS capsular types have been identified: types Ia, Ib, II, III, IV, V, VI, VII, VIII, and IX.

Epidemiology

GBS emerged as a prominent neonatal pathogen in the late 1960s. For the next 2 decades, the incidence of neonatal GBS disease remained fairly constant, affecting 1.0-5.4/1,000 liveborn infants in the USA. Two patterns of disease were seen: early-onset disease, which presents at <7 days of age, and late-onset disease, which presents at 7 days of age or later. In the 1990s, widespread implementation of maternal chemoprophylaxis led to a striking 65% decrease in the incidence of early-onset neonatal GBS disease in the USA, from 1.7/1,000 live births to 0.6/1,000 live births, whereas the incidence of late-onset disease remained essentially stable at approximately 0.4/1,000 (Fig. 177-1). Release of revised guidelines in 2002 coincided with a further reduction in the incidence of early-onset neonatal disease. In other developed countries, rates of neonatal GBS disease are similar to those in the USA prior to use of GBS chemoprophylaxis. In the developing world, GBS is not a major cause of neonatal sepsis, even though the prevalence of maternal vaginal colonization with GBS (a major risk factor for neonatal disease) among women from developing countries is similar to that reported among women living in the USA. The incidence of neonatal GBS disease is higher in premature and low birthweight infants, although most cases occur in full-term infants.

Figure 177-1 Incidence of early-onset and late-onset invasive group B streptococcus disease in three active surveillance areas (California, Georgia, and Tennessee), 1989 through 2000, and activities for the prevention of group B streptococcus disease. Arrows designate the dates when prevention activities occurred. ACOG, American College of Obstetricians and Gynecologists; AAP, American Academy of Pediatrics; CDC, Centers for Disease Control and Prevention.

(Adapted from Centers for Disease Control and Prevention: Early-onset group B streptococcal disease—United States, 1998-1999, MMWR 49:793–796, 2000; Schrag SJ, Zywicki S, Farley MM, et al: Group B streptococcal disease in the era of intrapartum antibiotic prophylaxis, N Engl J Med 342:15–20, 2000.)

Colonization by GBS in healthy adults is common. Vaginal or rectal colonization occurs in up to approximately 30% of pregnant women and is the usual source for GBS transmission to newborn infants. In the absence of maternal chemoprophylaxis, approximately 50% of infants born to colonized women acquire GBS colonization, and 1-2% of these infants develop invasive disease. Heavy maternal colonization increases the risk for infant colonization and development of early-onset disease. Additional risk factors for early-onset disease include prolonged rupture of membranes, intrapartum fever, prematurity, maternal bacteriuria during pregnancy, or previous delivery of an infant who developed GBS disease. Risk factors for late-onset disease are less well defined. Whereas late-onset disease may follow vertical transmission, horizontal acquisition from nursery or other community sources has also been described.

GBS is also an important cause of invasive disease in adults. GBS may cause urinary tract infections, bacteremia, endometritis, chorioamnionitis, and wound infection in pregnant and parturient women. In nonpregnant adults, especially those with underlying medical conditions such as diabetes mellitus, cirrhosis, or malignancy, GBS may cause serious infections such as bacteremia, skin and soft tissue infections, endocarditis, pneumonia, and meningitis. In the era of maternal chemoprophylaxis, most invasive GBS infections occur in nonpregnant adults. Unlike neonatal disease, the incidence of invasive GBS disease in adults has increased substantially, doubling between 1990 and 2007.

The serotypes most commonly associated with neonatal GBS disease are types Ia, III, V, Ib, and II. Strains of serotype III are isolated in more than 50% of cases of late-onset disease and of meningitis associated with early- or late-onset disease. The serotype distribution of colonizing and invasive strains from pregnant women is similar to that from infected newborns. In Japan, serotypes VI and VIII have been reported as common maternal colonizing serotypes, and case reports indicate that type VIII strains may cause neonatal disease indistinguishable from that caused by other serotypes.

Pathogenesis

A major risk factor for the development of early-onset neonatal GBS infection is maternal vaginal or rectal colonization by GBS. Infants acquire GBS during passage through the birth canal or, in some cases, via ascending infection. Fetal aspiration of infected amniotic fluid may occur. The incidence of early-onset GBS infection increases with the length of rupture of membranes. Infection may also occur through seemingly intact membranes. In cases of late-onset infection, GBS may be vertically transmitted or acquired later from maternal or nonmaternal sources.

Several bacterial factors are implicated in the pathophysiology of invasive GBS disease. Foremost among these is the type-specific capsular polysaccharide. Strains that are associated with invasive disease in humans elaborate more capsular polysaccharide than do colonizing isolates. All GBS capsular polysaccharides are high molecular weight polymers and contain a short side chain terminating in N-acetylneuraminic acid (sialic acid). Studies in type III GBS show that the sialic acid component of the capsular polysaccharide prevents activation of the alternative complement pathway in the absence of type-specific antibody. Thus, the capsular polysaccharide appears to exert a virulence effect by protecting the organism from opsonophagocytosis in the nonimmune host. In addition, type-specific virulence attributes are suggested by the fact that type III strains are implicated in most cases of late-onset neonatal GBS disease and meningitis. Type III strains are taken up by brain endothelial cells more efficiently in vitro than are strains of other serotypes, although studies using acapsular mutant strains demonstrate that it is not the capsule itself that facilitates cellular invasion. Other putative GBS virulence factors include GBS surface proteins, which may play a role in adhesion to host cells; C5a peptidase, which is postulated to inhibit the recruitment of polymorphonuclear cells into sites of infection; β-hemolysin, which has been associated with cell injury in vitro studies; and hyaluronidase, which has been postulated to act as a spreading factor in host tissues.

In a classic study, among pregnant women colonized with type III GBS, those who gave birth to healthy infants had higher levels of capsular polysaccharide-specific antibody than those who gave birth to infants who developed invasive disease. In addition, there is a high correlation of antibody to GBS type III in mother-infant paired sera. These observations indicate that transplacental transfer of maternal antibody is critically involved in neonatal immunity to GBS. Optimal immunity to GBS also requires an intact complement system. The classic complement pathway is an important component of GBS immunity in the absence of specific antibody; antibody-mediated opsonophagocytosis may also proceed via the alternative complement pathway. These and other results indicate that anticapsular antibody can overcome the prevention of C3 deposition on the bacterial surface by the sialic acid component of the type III capsule.

The precise steps between GBS colonization and invasive disease remain unclear. In vitro studies showing GBS entry of alveolar epithelium and pulmonary vasculature endothelial cells suggest that GBS may gain access to the bloodstream via invasion from the alveolar space, perhaps following intrapartum aspiration of infected fluid. β-Hemolysin/cytolysin may facilitate GBS entry into the bloodstream following inoculation into the lungs. However, highly encapsulated GBS strains enter eukaryotic cells poorly in vitro compared with capsule-deficient organisms, yet are associated with virulence clinically and in experimental infection models.

GBS induces the release of proinflammatory cytokines. The group B antigen and the peptidoglycan component of the GBS cell wall are potent inducers of tumor necrosis factor-α release in vitro, whereas purified type III capsular polysaccharide is not. Even though the capsule plays a central role in virulence through avoidance of immune clearance, the capsule does not directly contribute to cytokine release and the resultant inflammatory response.

The complete genome sequences of type III, V, and Ia GBS strains have been reported, emphasizing a genomic approach to better understanding GBS. Analysis of these sequences shows that GBS is closely related to Streptococcus pyogenes and Streptococcus pneumoniae. Many known and putative GBS virulence genes are clustered in pathogenicity islands that also contain mobile genetic elements, suggesting that interspecies acquisition of genetic material plays an important role in genetic diversity.

Clinical Manifestations

Two syndromes of neonatal GBS disease are distinguishable on the basis of age at presentation, epidemiologic characteristics, and clinical features (Table 177-1). Early-onset neonatal GBS disease presents within the 1st 6 days of life and is often associated with maternal obstetric complications, including chorioamnionitis, prolonged rupture of membranes, and premature labor. Infants may appear ill at the time of delivery, and most infants become ill within the 1st 24 hr of birth. In utero infection may result in septic abortion. The most common manifestations of early-onset GBS disease are sepsis (50%), pneumonia (30%), and meningitis (15%). Asymptomatic bacteremia is uncommon but can occur. In symptomatic patients, nonspecific signs such as hypothermia or fever, irritability, lethargy, apnea, and bradycardia may be present. Respiratory signs are prominent regardless of the presence of pneumonia and include cyanosis, apnea, tachypnea, grunting, flaring, and retractions. A fulminant course with hemodynamic abnormalities, including tachycardia, acidosis, and shock, may ensue. Persistent fetal circulation may develop. Clinically and radiographically, pneumonia associated with early-onset GBS disease is difficult to distinguish from respiratory distress syndrome. Patients with meningitis often present with nonspecific findings, as described for sepsis or pneumonia, with more specific signs of central nervous system (CNS) involvement initially being absent.

Table 177-1 CHARACTERISTICS OF EARLY- AND LATE-ONSET GBS DISEASE

Adapted from Schrag SJ, Zywicki S, Farley MM, et al: Group B streptococcal disease in the era of intrapartum antibiotic prophylaxis, N Engl J Med 342:15–20, 2000.

Late-onset neonatal GBS disease occurs on or after 7 days of life and most commonly manifests as bacteremia (45-60%) and meningitis (25-35%). Focal infections involving bone and joints, skin and soft tissue, the urinary tract, or lungs have been reported in approximately 20% of patients with late-onset disease. Cellulitis and adenitis are often localized to the submandibular or parotid regions. In contrast to early-onset disease, maternal obstetric complications are not risk factors for the development of late-onset GBS disease. Infants with late-onset disease are often less severely ill on presentation than infants with early-onset disease and the disease is often less fulminant.

Invasive GBS disease in children beyond early infancy is uncommon. In a multistate surveillance study in the 1990s, 2% of all cases of invasive GBS disease were identified in children age 90 days to 14 yr. Two of the more common syndromes associated with childhood GBS disease beyond early infancy are bacteremia and endocarditis. HIV infection should be considered in children with invasive GBS disease beyond the neonatal period.

Diagnosis

A major challenge is distinguishing between respiratory distress syndrome and invasive neonatal GBS infection in preterm infants because the 2 illnesses share clinical and radiographic features. Severe apnea, early onset of shock, abnormalities in the peripheral leukocyte count, and greater lung compliance may be more likely in infants with GBS disease. Other neonatal pathogens, including Escherichia coli and Listeria monocytogenes, may cause illness that is clinically indistinguishable from that due to GBS.

The diagnosis of invasive GBS disease is established by isolation and identification of the organism from a normally sterile site, such as blood, urine, or cerebrospinal fluid (CSF). Isolation of GBS from gastric or tracheal aspirates or from skin or mucous membranes indicates colonization and is not diagnostic of invasive disease. CSF should be examined in all neonates suspected of having sepsis, because specific CNS signs are often absent in the presence of meningitis, especially in early-onset disease. Antigen detection methods that use group B polysaccharide-specific antiserum, such as latex particle agglutination, are available for testing of urine, blood, and CSF, but these tests are less sensitive than culture. Moreover, antigen is often detected in urine samples collected by bag from otherwise healthy neonates who are colonized with GBS on the perineum or rectum.

Laboratory Findings

Frequently present are abnormalities in the peripheral white blood cell count, including an increased or decreased absolute neutrophil count, an elevated band count, an elevated ratio of bands to total neutrophils, or leukopenia. Elevations in the C-reactive protein level have been investigated as a potential early marker of GBS sepsis but are unreliable. Findings on chest radiograph are often indistinguishable from those of respiratory distress syndrome and may include reticulogranular patterns, patchy infiltrates, generalized opacification, pleural effusions, or increased interstitial markings.

Treatment

Penicillin G is the treatment of choice of confirmed GBS infection. Initial empirical therapy of neonatal sepsis should include ampicillin and an aminoglycoside (or cefotaxime), both for the need for broad coverage pending organism identification and for synergistic bactericidal activity. Once GBS has been definitively identified and a good clinical response has occurred, therapy may be completed with penicillin alone. Especially in cases of meningitis, high doses of penicillin (450,000-500,000 U/kg/day) or ampicillin (300 mg/kg/day) are recommended because of the relatively high mean inhibitory concentration of penicillin for GBS as well as the potential for a high initial CSF inoculum. The duration of therapy varies according to the site of infection (Table 177-2) and should be guided by clinical circumstances. Extremely ill near-term patients with respiratory failure have been successfully treated with extracorporeal membrane oxygenation.

Table 177-2 RECOMMENDED DURATION OF THERAPY FOR MANIFESTATIONS OF GBS DISEASE

Adapted from the American Academy of Pediatrics: Group B streptococcal infections. In Pickering LK, editor: Red book: 2000 report of the Committee on Infectious Diseases, ed 25, Elk Grove Village, IL, 2000, American Academy of Pediatrics, pp 537–544.

In cases of GBS meningitis, some experts recommend that additional CSF be sampled at 24-48 hr to determine whether sterility has been achieved. Persistent GBS growth may indicate an unsuspected intracranial focus or an insufficient antibiotic dose.

For recurrent neonatal GBS disease, standard intravenous antibiotic therapy followed by attempted eradication of GBS mucosal colonization has been suggested. This suggestion is based on the findings in several studies that invasive isolates from recurrent episodes are often identical to each other and to colonizing strains from the affected infant. Rifampin has most frequently been used for this purpose, but one report demonstrates that eradication of GBS colonization in infants is not reliably achieved by rifampin therapy. Optimal management of this uncommon situation remains unclear.

Prognosis

Studies from the 1970s and 1980s showed that up to 30% of infants surviving GBS meningitis had major long-term neurologic sequelae, including developmental delay, spastic quadriplegia, microcephaly, seizure disorder, cortical blindness, or deafness; less severe neurologic complications may be present in other survivors. Periventricular leukomalacia and severe developmental delay may result from GBS disease and accompanying shock in premature infants, even in the absence of meningitis. The outcome of focal GBS infections outside of the CNS, such as bone or soft tissue infections, is generally favorable.

In the 1990s, the case fatality rates associated with early- and late-onset neonatal GBS disease were 4.7% and 2.8%, respectively. Mortality is higher in premature infants; 1 study reported a case fatality rate of 30% in infants whose gestational age was <33 wk and 2% in infants whose gestational age was 37 wk or older. The case fatality rate in children aged 3 mo to 14 yr was 9%, and in nonpregnant adults was 11.5%.

Prevention

Persistent morbidity and mortality from perinatal GBS disease despite advances in neonatal care has spurred intense investigation into modes of prevention. Two basic approaches to GBS prevention have been investigated: (1) elimination of colonization from the mother or infant (chemoprophylaxis), and (2) induction of protective immunity (immunoprophylaxis).