Up to 1 million children, mostly in developing countries, die annually from infections caused by Streptococcus pneumoniae.1 Although pneumococcus is a well-documented cause of acute otitis media, it also contributes substantially to a variety of invasive disease processes that produce significant morbidity and mortality. For more than 20 years, rising antibiotic resistance rates have made treatment of invasive disease more difficult and have prompted the development of more inclusive vaccine products.
Current pneumococcal vaccines — the 23-valent polysaccharide vaccine (PPSV-23) intended for high-risk individuals older than 2 years available since the 1970s and the 7-valent conjugate vaccine (PCV7) introduced in 2000 for all infants — have proven effective in reducing rates of serious pneumococcal disease. The new 13-valent pneumococcal conjugate vaccine, (PCV13) granted approval by the U.S. Food and Drug Administration (FDA) in February, is currently being used and has already replaced PCV7 in hopes of curtailing rising rates of invasive disease attributed to some of the more prevalent non-PCV7 serotypes.
Nasopharyngeal carriage of S. pneumoniae is common in healthy children with its highest rates in infants; cumulative acquisition exceeds 90% before 6 months.2 Several factors have been shown to be associated with higher carriage rates — absence of breast-feeding, overcrowded living conditions, exposure to parental smoking, winter, and out-of-home child care attendance.2 Children younger than 2 years have the highest age-specific attack rates of invasive pneumococcal disease (IPD), peaking between 6 and 11 months. This same age group harbors pneumococcal isolates that are more likely to have antibiotic resistance when compared with older children and adults.3
Since the introduction of PCV7, there has been a profound decrease in the number of IPD cases caused by vaccine-containing serotypes, suggesting direct and herd effects of the vaccine.4,5 Unfortunately, serotype replacement in pneumococcal carriage, 6,7 as well as IPD with nonvaccine serotypes, have been reported in all age groups (see Figure 1). Additionally, resistance to commonly prescribed antimicrobials during the same time period has increased among the nonvaccine serotypes.8
Figure 1. The Trend of Invasive Pneumococcal Disease (IPD) According to the U.S. Pediatric Multicenter Pneumococcal Surveillance Group Since the Introduction of the 7-Valent Pneumococcal Conjugated Vaccine in 2007.
Pathogenesis and Immunity
There are at least 90 S. pneumoniae serotypes based on the capsular polysaccharide that surrounds the cell wall. In the widely accepted Danish system of nomenclature, serotypes are grouped according to antigenic similarities of the polysaccharide. The polysaccharide capsule is known to be the major virulence determinant, as it prevents phagocytosis by neutrophils and macrophages.9
Pneumolysin is a cytoplasmic protein manufactured by all pneumococci that promotes virulence in multiple ways, including pore formation (leading to cell lysis) and enhancement of nitric oxide production via macrophages (the latter strengthens antimicrobial defenses and contributes to host-induced tissue damage).10 The host’s innate defenses also play a role in the disease process, starting with mucosal barriers and surface enzymes. Type-specific anticapsular antibodies, made either in response to nasopharyngeal colonization by specific serotypes, or by vaccination with a pneumococcal-polysaccharide-containing vaccine, facilitate phagocytosis.
Invasive disease generally occurs within 1 month of nasal mucosal colonization with a serotype that the host has not been previously exposed to or immunized against, followed by an event, such as a viral respiratory infection or exposure to smoke, that predisposes the respiratory epithelium to pneumococcal invasion.9
Highest risk factors for IPD in U.S. children include immunodeficiency conditions (particularly HIV infection and hypogammaglobulinemia), splenic disorders, children with cochlear implants (meningitis), ethnic factors (African-American, Alaskan Native, American Indian), sickle cell disease, attendance at out-of-home childcare during the preceding 3 months, and age younger than 24 months.2 Children younger than 24 months are vulnerable because, following infection or vaccination with pure polysaccharide vaccines, they produce type-specific antibodies to only a few serotypes, such as type 3.2 In contrast, they do generate protective levels of antibody after vaccination with conjugated polysaccharide vaccines.
Pneumococcal disease in this population is rare and is generally thought to be associated with the maternal vaginal flora harboring S. pneumoniae.11 Although rare, invasive S. pneumoniae infections in the neonate (SPIN) are of significant concern because of their substantial morbidity and mortality (14% to 60%), as well as the lack of vaccine availability for this patient population.12,13
Neonatal IPD can manifest as bacteremia, early- and late-onset sepsis, pneumonia, septic arthritis, osteomyelitis, and meningitis, often accompanied by leukopenia/neutropenia. Although earlier studies report these presentations within the first few days of life,12 the U.S. Pediatric Multicenter Pneumococcal Surveillance Group (USPMPSG) demonstrated onset of disease usually at 2 to 3 weeks and found that 26% of SPIN were caused by serogroups 1, 3, 5, and 12, which are not in PCV7;13 types 1, 3, and 5 are in PCV13.
The method of pneumococcal acquisition in neonates is still undetermined because previous studies showed S. pneumoniae colonization of the maternal genital tract to be extremely rare (≤ 0.03%).14 This has led several authors to suggest a high invasion to colonization rate in neonates.13 Although considered to be relatively rare, SPIN should not be forgotten when evaluating ill neonates.
Bacteremia Without a Focus
Regardless of age, bacteremia without a focus (also known as occult bacteremia) is the most common form of IPD, though it is most often seen in children aged 3 to 24 months.15 In the post-Haemophilus influenzae type b vaccine/pre-PCV7 era, S. pneumoniae was the etiologic agent in 88% to 92% of occult bacteremia cases.2 Factors associated with an increased likelihood of pneumococcal bacteremia include fever without an obvious source, peak temperatures higher than 40°C, and an elevated peripheral white blood cell count (> 15,000/mm3).9 Only about 10% of children with occult bacteremia develop systemic complications.16
The musculoskeletal system can also fall victim to IPD, ranging from bone and joint infections to rhabdomyolysis. The femur and the humerus are the bones most often affected (although unusual sites — calcaneous, ileum, and rib — have also been reported), and the knee and hip are the most commonly involved joints.17
Although not as frequent a culprit of such infections when compared with Staphylococcus aureus or Streptococcus pyogenes, up to 4% of bacterial bone and up to 20% of bacterial joint infections in infants and children are caused by S. pneumoniae.17 Of concern are the younger patients who often present with fever without a focus — 10% of patients found to have musculoskeletal pneumococcal disease present in this fashion.17 This same population has higher rates of sequelae than their elder counterparts, suggesting that immunologic or anatomic factors in young infancy may predispose to more destructive musculoskeletal disease.17
Upper Respiratory Tract Infections Other than Otitis Media
Mastoiditis is the most common complication of acute or chronic otitis media within the temporal bone. Streptococcus pneumoniae is the most common cause of acute mastoiditis in children and is the most commonly isolated pathogen in patients with sinusitis. Pneumococcal mastoiditis occurs primarily in children younger than 2 years and has been seen more frequently in recent years, following an initial decrease in the number of cases as a result of the introduction of PCV7. These increased numbers are directly related to the increased prevalence of the 19A serotype.18 This particular serotype is not in PCV7 and is associated with significant antibiotic resistance. Sinusitis can occur at any age, commonly affects the ethmoid and maxillary sinuses, and initially manifests as a viral infection that is followed by a purulent nasal discharge and cough.9
Infective endocarditis is a rare event occurring in 0.4% of 3,065 children with pneumococcal bacteremia identified during a 9.5-year period by the USPMPSG. S. pneumoniae accounts for 3% to 7% of all pediatric endocarditis cases, and mostly occurs in patients with pre-existing structural heart disease.18 Patients may have concomitant pneumococcal disease affecting other systems (ie, meningitis, pneumonia, otitis media) and present with nonspecific complaints. A male predominance has been noted in multiple studies, while the median age has ranged from 15 months to 9 years.19 Many of the responsible serotypes previously reported are included in PCV7.
Central Nervous System
S. pneumoniae is the leading cause of bacterial meningitis in children and is associated with rising rates of antimicrobial resistance over the years. Pneumococcal meningitis is associated with significant morbidity, including neurologic sequelae (such as hearing loss and motor deficits), which occur in at least one-third of patients.20 Although signs and symptoms can vary, clinical presentations of meningitis include nuchal rigidity, lethargy, and fever.
The Active Bacterial Core surveillance of the Center for Disease Control and Prevention (CDC) Emerging Infections Program network reported a dramatic decline in pneumococcal meningitis from 1998 to 2005 (attributable to the introduction of PCV7 in 2000) — a 64% decrease in incidence among children younger than 2 years.21 In fact, rates of adult pneumococcal meningitis have also decreased since the introduction of PCV7.21,22
Although Hsu et al.21 demonstrated fewer penicillin-nonsusceptible strains from 1998 to 2003, a recent increase in penicillin-resistant pneumococcal meningitis is concerning. Fortunately, studies have not shown a significant difference in clinical outcomes related to the penicillin or ceftriaxone susceptibility of the causative S. pneumoniae.20 Although the data in relation to dexamethasone use in pneumococcal meningitis are not as conclusive as has been reported for the initial treatment for Haemophilus influenzae type b meningitis, many experts advocate it concomitantly with antimicrobials. However, according to their 2009 Red Book (page 528), the Committee on Infectious Diseases of the American Academy of Pediatrics makes no recommendations for its use in children.
Lower Respiratory Tract
Pneumonia is the leading cause of death and disease worldwide in terms of IPD.1 Although pneumococcus can reside asymptomatically on the upper respiratory mucosal surface, access to the normally sterile aspects of the lower respiratory tract can result in pneumonia, making S. pneumoniae one of the most common bacterial causes of communityacquired pneumonia.23
The “classic” presentation of pneumococcal pneumonia that occurs in older children is described as a prodromal viral-like respiratory illness, followed by abrupt onset of shaking chills, high fever, pleural pain, dyspnea, prostration, rust-colored sputum production, and findings (physical/radiographic) suggestive of lobar consolidation.2
Otherwise, clinical manifestations can vary widely and can be very non-specific, especially in the young infant. Fever, vomiting, abdominal distension and pain mimicking an acute abdomen, and nuchal rigidity mimicking meningitis are seen occasionally. In infants and young children, a bronchopneumonia radiographic pattern is common.
Since the implementation of PCV7 in infants and children, active populationand laboratory-based surveillance by the CDC has demonstrated significant reductions in ambulatory-care visits for allcause pneumonia and the incidence and hospitalization rates of all-cause pneumonia in children younger than 2 years.24 However, the same CDC surveillance found that PCV7 had not altered the rate for all-cause pneumonia among children 2 to 4 years. Tan et al.25 showed that penicillin nonsusceptible pneumococci were no more likely to cause complicated pneumonia than susceptible strains, although the numbers in the study were limited.
Although many episodes of community-acquired pneumonia can be treated as an outpatient, numerous complications associated with S. pneumoniae can contribute to the morbidity seen with pneumonia — parapneumonic effusions, empyema, lung necrosis, bronchopleural fistulas, and abscesses (see Figure 2). Furthermore, Li et al.26 found an increase in pediatric empyema hospitalizations from 1997 through 2006 despite decreases in bacterial pneumonia and IPD rates that are attributed to immunization with PCV7. Data in Utah revealed an increase in pneumococcal necrotizing pneumonia since the utilization of PCV7, most commonly caused by serotype 3, a nonvaccine serotype, which is included in PCV13.27 These complications often require further treatment with invasive procedures, including thoracentesis, chest tube placement, thoracoscopy and lobectomy, which further contribute to the morbidity seen in pneumococcal disease.
Figure 2. Pulmonary CT Scan in a 3-Year-Old Child with Pneumococcal Necrotizing Pneumonia, Empyema, and a Bronchopleural Fistula Resulting in a Pyopneumothorax.
Skin and Soft Tissue
S. pneumoniae is an uncommon cause of skin and soft tissue infections (SSTI), although it can be seen more frequently in patients with some degree of immunodeficiency, particularly in HIV infection, connective tissue disease, and prolonged corticosteroid therapy.28 Givner et al.29 described 52 cases of pneumococcal facial cellulitis (45 periorbital and seven buccal) with associated bacteremia in children, of whom 92% were younger than 36 months and 88% were previously healthy. One of the largest series of S. pneumoniae SSTI is from Spain, where 2.2% of all pneumococcal isolates were from skin and soft tissue samples (69 patients), including surgical wound infections, burn infections, and pyomyositis.30 Other soft tissue infections reported include erysipelas, glossitis and cystic gingival lesions.2
Although usually seen in patients with underlying conditions, such as nephrotic syndrome or immunocompromised status, primary (without a concurrent intraabdominal process) pneumococcal peritonitis should be considered in the differential diagnosis of children with an acute abdomen.9 Pneumococcal peritonitis can often mimic acute appendicitis, especially in the female population, in whom pneumococcal peritonitis is most often diagnosed.
Hemolytic uremic syndrome (HUS) caused by IPD is associated with significant morbidity and mortality. However, it has been considered underrecognized by experts, likely caused by the absence of specific laboratory testing and consistent case definitions, as well as unfamiliarity with this association.31
HUS occurs when pneumococcal neuraminidase unmasks the Thomsen-Friedenreich antigen (T antigen) located on the cell surfaces of erythrocytes, platelets, and glomeruli. Naturally occurring immunoglobulin M antibodies bind to the exposed T antigen and produce the manifestations of HUS — microangiopathic anemia, renal injury, and thrombocytopenia.
HUS associated with IPD tends to occur in younger infants, patients who have received unwashed blood products, and in those diagnosed with meningitis or complicated pneumonia. Since introduction of PCV7, 19A has become the most common serotype seen in S. pneumoniae-associated HUS.31 It has been postulated that PCV7 has selected for this serotype, which may possess increased neuraminidase activity.31
Children with sickle-cell disease, congenital asplenia, or previous splenectomy can experience a rapidly progressive, fulminant course of sepsis resembling the Waterhouse-Friderichsen syndrome. Patients who suffer from asplenia are unable to properly filter blood-borne bacteria and coordinate an appropriate immune attack by various cell types, providing a perfect environment for IPD that tends to lead to significant mortality.
HIV-infected children have a higher rate of pneumonia and septic shock than their non-HIV-infected counterparts.32 A 13-year prospective study from the USPMPSG found that 80% of children with recurrent systemic pneumococcal disease had an underlying disease, most commonly HIV.33 The same group reviewed IPD in bone marrow (BMT) and solid organ transplant (SOT) patients.34 Of 51 episodes of IPD, 19 occurred in BMT and 24 in SOT patients. Eight patients (two BMT and six SOT) had multiple episodes. Only 33% of eligible patients had received PPSV-23.
Antibiotic Therapy and Resistance
Penicillin G remains the drug of choice for susceptible pneumococci, even for severe infections in which clinicians may prefer a more broad-spectrum beta-lactam. Unfortunately, there has been an increase in penicillin-nonsusceptible strains of S. pneumoniae, with the highest rates found in infants.
Prospective surveillance from September 1993 through August 1999 by the USPMPSG revealed a fourfold increase in penicillin-resistance (4% to 15%) and ceftriaxone-resistance rates (0.5% to 2%) for IPD; the most common serotypes/serogroups associated with penicillin-nonsusceptibility are included in PCV7.35 Although macrolides, clindamycin, and trimethoprimsulfamethoxazole are alternatives to beta-lactams for therapy, increasing resistance among penicillin-nonsusceptible strains limits their use.
When pneumococcal meningitis is suspected, empiric therapy should consist of vancomycin (15 mg/kg/dose every 6 hours) and ceftriaxone (100 mg/kg/day divided every 12 or 24 hours) or cefotaxime (100 mg/kg/dose every 8 hours). Once susceptibility testing is available, therapy should be simplified based on the results. For IPD outside of the central nervous system, the USPMPSG found retrospectively that ceftriaxone, cefotaxime, or cefuroxime are adequate agents for pneumococcal isolates with minimum inhibitory concentrations ≤ 2 μg/ml.36
The use of antibiotics to prevent IPD is focused on patients with anatomic or functional asplenia and immunodeficiency conditions. Penicillin has been the most studied antimicrobial in this patient population. Use is approved up to at least 5 years, although many providers prefer coverage through adulthood, especially in sickle-cell patients.
Since the 1980s, a 23-valent polysaccharide vaccine has been available for use in patients younger than 2 years of age and includes antigens from 23 serotypes that cause approximately 90% of IPD in adults and children.9
Unfortunately, IPD also affects children younger than 2 years who cannot benefit from immunization with this pure polysaccharide vaccine because of the inability to mount the necessary T-cell-independent immune response.9 This type of immune response produces protective antibodies and induces immunologic memory, which allows a boosting effect with subsequent doses of vaccine. In 2000, PCV7 containing serotypes 4, 6B, 9V, 14, 18C, 19F, and 23F was introduced and has been shown to significantly reduce the rates of IPD in children, including those younger than 2 years who are able to establish a T-dependent response, modestly decrease rates of noninvasive pneumococcal disease, such as acute otitis media, and reduce the rates of IPD in adults and immunocompromised individuals through herd immunity.37
However, concerns exist over serotype replacement and antibiotic resistance. A recent study investigating nasopharyngeal S. pneumoniae carriage in Massachusetts children during the two time periods 2000 to 2001 and 2006 to 2007 noted an increase of non-PCV7 serotypes from 15% to 29% and a significant increase in penicillin nonsusceptibility among the non-PCV7 serotypes identified.38
The USPMPSG found that since 2005, the number of invasive pneumococcal infections in children has increased at eight pediatric institutes — most commonly presenting as pneumonia. Additionally, more than 95% of IPD was caused by nonvaccine strains, primarily serotype 19A, which also demonstrated a dramatic rise in penicillin and ceftriaxone resistance.7
In February, the FDA approved a 13-valent conjugated pneumococcal vaccine which contains serotypes 1, 3, 5, 6A, 7F, and 19A, in addition to those in PCV7 for use in children 6 weeks to 5 years of age.38 Based on 2007 data, more than 60% of IPD caused by nonvaccine serotypes were caused by one of the six additional serotypes included in the PCV13. PCV13 has been compared with PCV7 for immunogenicity and safety.39 Serotype-specific antipolysaccharide IgG antibody and functional opsonophagocytic assay responses suggest that PCV13 is at least as immunogenic as PCV7 for the seven common serotypes and, as expected, much more immunogenic for the additional six serotypes. PCV13’s safety and tolerability profile was similar to that of PCV7.
The infant immunization schedule for the PCV13 is the same as that used for PCV7 — four doses at 2, 4, 6, and 12 to 15 months. Children who have received one or more doses of PCV7 may complete the four-dose immunization series with PCV13. If a child has been fully immunized with four doses of PCV7, the Advisory Committee on Immunization Practices of the CDC recommends that healthy children up to 59 months receive one dose of PCV13 to elicit antibodies against the six additional serotypes with a minimum of 8 weeks since the last PCV7. Although FDA-licensed for use in children from 6 weeks through 71 months of age, PCV13 may be administered to children 6 through 18 years of age with sickle-cell disease, HIV, or other immunocompromised condition.