Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Group B streptococcus and Escherichia coli account for most of the cases of neonatal early-onset bacterial sepsis.
Prevention of infection by maternal treatment is the main factor accounting for the decreased incidence of early-onset group B streptococcus sepsis but does not affect rates of late-onset group B streptococcus sepsis.
Microbiologic blood cultures of adequate volume (at least 1.0 mL) represent the mainstay for the diagnosis of infection.
Management of the asymptomatic newborn at risk of infection due to maternal risk factors is evolving, with evidence supporting evaluation and treatment of only a small proportion of asymptomatic infants.
Ampicillin and gentamicin are recommended as initial therapy in neonates with suspected bacterial sepsis. Treatment can then be narrowed appropriately after the results of antibiotic susceptibility studies.
Neonatal sepsis is one of the leading causes of morbidity and mortality of newborns worldwide. It is estimated that nearly 430,000 neonatal deaths occurred in 2013 due to neonatal sepsis, approximately 15% of all neonatal deaths. Thus, neonatal sepsis is a common and deadly disease process that requires an understanding of the pathogenesis and timely treatment to prevent severe morbidity or mortality.
Bacterial sepsis in newborns is commonly divided into two categories: early-onset sepsis (EOS) and late-onset sepsis (LOS) based upon time since birth. Traditionally, EOS is classified as either occurring in the first 72 hours or, in some cases, within the first 7 days after birth. In this chapter, we will define EOS as occurring within the first 72 hours after birth and LOS as occurring after 72 hours. Here, we will discuss early-onset neonatal bacterial infections, including the pathogenesis, epidemiology, clinical signs and symptoms, common bacterial pathogens, evaluation methodology, and prevention strategies. We will then explore late-onset neonatal bacterial infections and conclude with an evaluation of the medical literature surrounding neonatal bacterial meningitis.
Early-onset bacterial sepsis remains a major cause of neonatal morbidity and death, although the sepsis-associated death rates per 100,000 live births has declined significantly from 2000 to 2015. The introduction of intrapartum antibiotic prophylaxis (IAP) in pregnant women during labor and delivery contributed to much of this decline in mortality. Mortality rates in infected premature and very immature neonates are significantly higher than those in term neonates. Significant improvements in neonatal intensive care and early identification of infected neonates have contributed to reduced mortality rates in the newborn period.
Vertical transmission is the main route through which EOS occurs and involves the passage of pathogenic bacteria from the mother to the neonate before or during delivery. Maternal chorioamnionitis and acute inflammation from presumed infection of the membranes surrounding the fetus in utero is a well-known risk factor for EOS. Other potential routes for group B streptococcus (GBS) acquisition and EOS include ascending spread from the vaginal environment to the uterus following membrane rupture or fetal contact with pathogenic bacteria during passage through the birth canal at delivery. Less commonly, bacteria reach the fetus through the in utero transplacental passage, as suggested by the presence of high-grade bacteremia and severe sepsis that are clinically apparent at the time of birth in the presence of an intact membrane in neonates born via cesarean delivery. Although organisms recovered from the amniotic sac in the mother are usually polymicrobial and include organisms such as GBS, group D enterococcus, aerobic gram-negative bacteria, and anaerobes such as Bacteroides spp., a single organism causing bacterial sepsis is the rule in sepsis of the newborn.
Pathogenesis of neonatal sepsis is due to several key factors, including but not limited to timing and duration of exposure, inoculum size, the virulence of the pathogenic bacteria, and the immune status of the neonate. Although many microorganisms recovered from the amniotic cavity are thought to induce spontaneous preterm labor and possibly premature rupture of membranes, the exact mechanisms by which this may occur are debatable. Clinical or subclinical chorioamnionitis can incite a marked inflammatory response with the release of cytokines that can contribute to the onset of preterm labor and premature rupture of membranes. Other risk factors for clinical intra-amniotic infection include young maternal age, prolonged labor, prolonged rupture of membranes (≥18 hours), internal scalp fetal monitoring, the presence of urinary tract infections (UTIs), and a history of bacterial vaginosis.
The incidence of early-onset bacterial infection is variable and ranges from 1 to 5 per 1000 live births. Rates of neonatal sepsis are inversely related to gestational age. Notably, there has been a reduction in the rate of EOS in the United States since the implementation of IAP against GBS. Early-onset GBS infection rates in the United States have declined from 0.37 per 1000 live births in 2006 to 0.23 per 1000 live births in 2015. However, despite IAP, late-onset rates have remained relatively stable at approximately 0.31 per 1000 live births.
Infants with EOS frequently have one or more identifiable risk factors. Prematurity is the single greatest risk factor for EOS. Because extremely low birth weight infants have impaired host defenses and since preterm birth may be associated with low-grade chorioamnionitis, it is not surprising that premature neonates have nearly 30 times higher rates of mortality from EOS compared to term neonates (1.6% in term neonates vs. 30% mortality in those at 25 to 28 weeks of gestation). Neonatal susceptibility to GBS infection is increased with deficiencies in circulating levels of GBS type-specific antibody and complement, and is further heightened by neutrophil dysfunction seen in infants who are more premature. Other risk factors for EOS are maternal age, health and nutrition, colonization with well-known pathogens (e.g., GBS), maternal fever, and longer duration of rupture of membranes.
The most common bacterial pathogens associated with EOS are GBS and Escherichia coli . Here we review these two primary pathogens in depth before briefly exploring other less common bacterial etiologies of EOS, such as Listeria monocytogenes and other gram-negative enteric bacilli.
GBS accounts for nearly 45% of culture-confirmed cases of EOS in term neonates and 25% among very low birth weight (VLBW) infants (weighing <1500 g at birth). Approximately 20% to 30% of pregnant women in the United States are colonized with GBS. Treatment of GBS-colonized women during pregnancy only temporarily eradicates the organism. Most women are recolonized within several weeks. Before the introduction of IAP, approximately 50% of neonates born to mothers known to be GBS carriers became colonized, and approximately 1% to 2% of colonized neonates developed GBS infection. Since the widespread use of IAP, approximately 60% to 80% of GBS cases occur in infants born to mothers who screened negative for GBS.
Detection of maternal GBS colonization has been emphasized since 1996, approximately. Early studies determined that the optimal sampling sites for GBS were the genitourinary and gastrointestinal tracts. The optimal time for performing prenatal cultures is between 36 0/7 and 37 6/7 weeks' gestation, and the highest culture yield is obtained when both the lower vaginal area and anal or rectal sites are sampled. Furthermore, GBS testing should be performed on pregnant women who present in preterm labor or with the rupture of membranes (PROM) before 37 0/7 weeks’ gestation. If GBS has previously been detected through urine culture during pregnancy, no further confirmation by vaginal-rectal culture is necessary. By identifying the pregnant woman who is colonized with GBS, IAP can be provided in a timely manner. Since the introduction of IAP recommendations, the national incidence of GBS EOS has decreased from 1.8 cases (1990) to 0.23 cases (2015) per 1000 live births. However, late-onset GBS incidence has not changed.
To prevent GBS EOS, IAP is recommended for all women with GBS colonization identified by antenatal recto-vaginal culture, with a history of GBS bacteriuria during the same pregnancy, with a history of a former infant with GBS disease, or for women who present in preterm labor or have PROM <37 0/7 weeks’ gestation. For women who present in labor ≥37 0/7 weeks’ gestation with unknown GBS status should have IAP if maternal temperature ≥38°C or duration of rupture of membranes ≥18 hours.
GBS is sensitive to penicillin, which is the drug of choice because of its narrow spectrum; the alternative is ampicillin. If a mother is allergic to penicillin but considered low-risk for anaphylaxis, the use of cefazolin is recommended. When patients are at a high risk of anaphylaxis, tests for antimicrobial susceptibility of prenatal GBS to clindamycin should be performed. In 2016, 42% of GBS isolates were found to be resistant to clindamycin. Therefore if the prenatal GBS is sensitive to clindamycin, patients can receive clindamycin intravenously (IV) until delivery. Of note, erythromycin is no longer recommended as an IAP due to high resistance patterns. When GBS is resistant to clindamycin, IV vancomycin every 12 hours until delivery is recommended. The high levels of resistance highlight the importance of antibiotic susceptibility testing on the group B streptococcal isolates from pregnant women. Only penicillin G, ampicillin, or cefazolin are considered adequate IAP. If clindamycin or vancomycin are used, while likely capable of providing some protection against invasive early-onset GBS disease in the newborn, it is considered inadequate IAP coverage due to limited data on their clinical efficacy.
There are 10 known serotypes of GBS (I, Ia, II-IX) that are distinguished by a specific polysaccharide capsule. The polysaccharide capsule helps the GBS evade complement deposition, opsonization, and phagocytosis by the infant’s immune system and is considered the most important virulence factor. Other virulence factors include proteases (such as C5a peptidase), which cleave complement and surface proteins (such as alpha and beta C-proteins) and promote evasion of human host defenses. Serotype III is associated with an estimated 62% of invasive infant strains and 25% of colonizing strains and is also a common cause of GBS meningitis. Globally, serotypes I to V are responsible for the vast majority, 98% of colonization and 97% of invasive infant strains. Multivalent vaccines against specific capsular polysaccharides have been investigated through phase I and II clinical trials and demonstrated their safety and immunogenicity. Further clinical trials are planned.
While most EOS is classified as sepsis in the first 72 hours after birth, for GBS infection, early-onset disease (EOD) is classified in the first 7 days after birth. Subsequently, late-onset disease occurs from 1 week to 3 months of age, and very-late-onset disease occurs more than 3 months after birth ( Table 33.1 ). Although IAP has led to a significant decrease in the incidence of GBS EOD, there is no evidence that chemoprophylaxis prevents late-onset or very-late-onset disease. GBS EOD is generally believed to be caused by ascending infection from the maternal birth canal to the uterine compartment, with subsequent fetal infection through colonization of the skin or aspiration of infected amniotic fluid. The strongest risk factor for GBS EOD is maternal GBS colonization. However, even if a mother has screened negative for GBS, GBS EOD can still occur due to maternal colonization status during the time from screening to delivery. Evaluation of neonates at risk for GBS EOD is discussed separately below.
Characteristic | Early Onset | Late Onset |
---|---|---|
Age at onset | Birth to 7 days of life | 7 days to 3 months of life |
Symptoms | Respiratory distress, apnea | Irritability, fever, poor feeding |
Findings | Pneumonia, sepsis | Sepsis, meningitis, osteoarthritis |
Maternal obstetric complications | Frequent | Uncommon |
Mode of transmission | Vertical, in utero, or intrapartum | Nosocomial, horizontal |
Predominant GBS serotypes | Ia, III, V * | III, Ia, V * |
Effect of intrapartum antibiotic prophylaxis recommended by the Centers for Disease Control and Prevention | Reduces incidence by 85%–90% | No effect |
Historically, E. coli is the second most common pathogen causing sepsis and meningitis in newborns. The antigenic structure of E. coli is complex and is composed of approximately 150 somatic or cell wall O antigens, 50 flagellar H antigens, and approximately 80 capsular K antigens. However, a limited number of specific K antigen E. coli strains cause meningitis. Approximately 80% of the strains causing meningitis and 40% of the strains causing bacteremia or sepsis express the K1 antigen. The capsular K1 polysaccharide antigen is highly homologous to the capsular antigen of group B Neisseria meningitidis . Because a high percentage of women have bacteriuria with strains of E. coli that express the K1 antigen or are colonized with it at the time of delivery, it is surprising that E. coli sepsis or meningitis is not more common.
Surveillance data from the National Institute of Child Health and Human Development Neonatal Research Network, a consortium of 16 US academic neonatal centers, revealed that in the era of widespread implementation of antibiotic prophylaxis, the rate of E. coli sepsis increased from 3.2 to 6.8 cases per 1000 live births. This increase was observed in the 1998–2000 era and persisted from 2002 to 2003. Approximately 85% of E. coli infections in VLBW infants were ampicillin-resistant. Yet, while most evidence suggests that IAP has not been associated with a concomitant increase in the incidence of E. coli or other non-GBS bacterial causes, other evidence demonstrates that the incidence of E. coli and ampicillin-resistant E. coli infections increased significantly among preterm infants. Furthermore, neonates who developed E. coli infections with ampicillin-resistant strains are more likely to be born from mothers with IAP with ampicillin. In VLBW neonates, the incidence of EOS with E. coli has increased, with nearly 85% of cases having resistance to ampicillin. Thus, while the benefits of IAP on reducing EOS attributable to GBS are well-documented, the balance of preventing resistance by other bacterial pathogens is still under investigation.
L. monocytogenes is a small, facultative anaerobic, gram-positive, motile bacillus that produces a narrow zone of beta hemolysis on blood agar plates. It can be confused with GBS unless a careful Gram stain, a catalase reaction, and other tests are performed. Most L. monocytogenes infections are due to three serotypes: 1a, 1b, and 4b. The last serotype has been described in most outbreaks of listeriosis.
Most cases of listeriosis appear to be food-borne. Foods commonly contaminated by L. monocytogenes include raw vegetables such as cabbage, raw milk products, fish, poultry, processed chicken, beef, and hot dogs. Transmission to the fetus occurs through either a hematogenous (transplacental) route or via an ascending infection through the birth canal. Frequently, infections with Listeria spp. early in gestation result in abortion; later in pregnancy, infection with Listeria spp. can result in the premature delivery of a stillborn or infected newborn. Approximately 70% of Listeria -infected women deliver before 35 weeks' gestation.
L. monocytogenes illness in the mother may be undetected as it can be a vague influenza-like illnesses that may not come to medical attention. In approximately half of the perinatal cases, illness in the mother has preceded delivery by 2 days to 2 weeks. An autopsy of stillborn neonates or those who die in the perinatal period from L. monocytogenes , granulomas are found throughout organs such as the liver and lungs, and infection is widely disseminated, including involvement of the meninges and even a skin rash due to microabscesses and granulomas called granulomatosis infantisepticum. Treatment of Listeria spp. infection or bacteremia during pregnancy can prevent infection in the fetus.
Like GBS infection, Listeria spp. infection may have either an early-onset or a late-onset presentation. Epidemics of neonatal Listeria spp. infection has been described after the ingestion of contaminated foods such as cheese or coleslaw. The first clearly documented food-borne (coleslaw) outbreak of listeriosis was in 1981 from the Maritimes in Canada; it was associated with a fatality rate of 27%. There are reports of repeated abortions in women with colonization of Listeria spp. in the gastrointestinal tract, and cultures can be performed to detect fecal carriage in such women; selective media for Listeria spp. is recommended to isolate the organism from various foods or stool specimens. Rapid antigen tests based on nucleic acid amplification are available but not commonly used in clinical diagnostic laboratories. There is no vaccine for Listeria spp. infection, but preventive measures have included the surveillance programs from the US Department of Agriculture, prohibiting the sale of contaminated meats. Between 1996 and 2006, the incidence of Listeria spp. infections declined by 36%; however, an outbreak of the disease in 2002 related to contaminated turkey meat led to 54 illnesses, 8 deaths, and 3 fetal deaths in 9 states within the US.
The bacteria responsible for early-onset neonatal sepsis have changed over time, and there are regional differences in the organisms commonly responsible for early-onset sepsis. In addition to the organisms mentioned previously, other bacterial pathogens associated with early-onset bacteremia or sepsis in newborns include Enterococcus spp., viridans group Streptococcus spp., Klebsiella spp., Enterobacter spp., Haemophilus influenzae (typeable and nontypeable), Staphylococcus aureus, Streptococcus pneumoniae , group A streptococcus and other beta-hemolytic streptococci, and coagulase-negative staphylococci (CONS).
There is great variability in the clinical presentations of neonates with early-onset bacterial sepsis ( Box 33.1 ). Most neonates exhibit respiratory distress in the first 12 hours of life, frequently immediately after birth. In these neonates, the progression may be rapid, with cardiovascular instability, shock, and death. Presentation within the first 12 hours of life suggests that the infection occurred at or near the time of birth or during the immediate postnatal period. Neonates with in utero hypoxia may gasp, inhaling contaminated amniotic fluid and setting the stage for early-onset pneumonia, bacteremia, and sepsis.
Abnormal neurologic status: irritability, lethargy, poor feeding.
Abnormal temperature: hyperthermia or hypothermia.
Apnea.
Bleeding problems: petechiae, purpura, oozing.
Cardiovascular compromise: tachycardia, hypotension, poor perfusion.
Cyanosis.
Gastrointestinal symptoms: abdominal distention, emesis, diarrhea.
Jaundice.
Respiratory distress: tachypnea, increased work of breathing, hypoxemia.
Seizures.
The signs of early-onset infection may be subtle, such as tachypnea, or may be more overt, with respiratory distress and hypotension. Because the signs of sepsis can be relatively nonspecific, such as poor feeding and increased sleepiness, they can be overlooked. The clinical signs of neonatal sepsis include, but are not limited to, temperature instability (hyperthermia or hypothermia); respiratory distress; cyanosis; jaundice; hepatomegaly; abdominal distention; feeding abnormalities; and neurologic abnormalities (including lethargy, apnea, and seizures).
Become a Clinical Tree membership for Full access and enjoy Unlimited articles
If you are a member. Log in here