Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Streptococcus pneumoniae
Long recognized for causing asymptomatic colonization and as a prominent cause of pneumonia, bacteremia, meningitis, sinusitis, and otitis media, Streptococcus pneumoniae is likely the most common cause of serious bacterial respiratory infection in both children and adults worldwide. Antimicrobials and vaccines have substantially reduced the incidence of, and morbid outcomes from, pneumococcal infection. However, acquisition of antibiotic resistance, the more limited impact of vaccines on mucosal disease (e.g., pneumonia, otitis media), the emergence of nonvaccine serotypes, and a growing immunocompromised population provide challenges for ongoing control of this prevalent and invasive pathogen.
History
S. pneumoniae has played a prominent role in the history of microbiology. Identified concurrently in 1881 in France by Pasteur and in the United States by Sternberg, this bacterium was soon recognized as the most common cause of lobar pneumonia and became known as the pneumococcus. Based on its appearance in Gram-stained sputum, the name Diplococcus pneumoniae was assigned in 1926 and changed to Streptococcus pneumoniae in 1974, based on its morphology during growth in liquid medium.
S. pneumoniae was the first organism to be recognized as showing characteristics of a prototypic extracellular bacterial pathogen by replicating extracellularly in mammalian tissues. Resistance to uptake by phagocytic cells is largely related to the polysaccharide capsule. In the early 1890s, Felix and Georg Klemperer showed that immunization with killed pneumococci protected rabbits against subsequent pneumococcal challenge and that protection could be transferred by infusing serum (“humoral” substance) from immunized rabbits into naïve recipients. The basis for this immunity was shown by Neufeld and Rimpau to be the presence of factor(s) in serum that facilitate ingestion by white blood cells (WBCs), a process they called opsonization, derived from the Greek word for buying provisions. These observations provided the basis for what we now call humoral immunity. Individual capsular serotypes were recognized when injection of killed organisms into a rabbit stimulated the production of serum antibody that agglutinated and caused capsular swelling (“Quellung”) of the immunizing strain, as well as some, but not all, other pneumococcal isolates. Early in the 20th century, serotypes 1, 2, and 3 were distinguished, and all other pneumococci were called group 4.
In the first decade of the 20th century, the concept of humoral immunity was used to address the problem of epidemic lobar pneumonia that each year affected as many as 1 in 10 African miners. A program of immunization with whole killed pneumococci substantially reduced the incidence of pneumonia. In the 1920s, Heidelberger and Avery demonstrated that the protective antibody reacted with surface capsular polysaccharides, the first nonprotein antigen identified. Felton prepared the first purified pneumococcal capsular polysaccharides (PPSs) for immunization of human subjects, and mass vaccination with Felton's type 1 polysaccharide aborted an epidemic of pneumonia at a state hospital in Massachusetts in the winter of 1937 to 1938. Thus specific bacterial polysaccharide antigens could be used to stimulate antibodies that conferred protection against epidemic human infection. These observations were confirmed during World War II when MacLeod and coworkers showed that vaccinating military recruits with capsular material from four serotypes of S. pneumoniae greatly reduced the incidence of pneumonia caused by serotypes in the vaccine but not by other serotypes.
S. pneumoniae also played a central role in the discovery of DNA. Experiments by Griffith in the 1920s revealed that intraperitoneal injection into mice of live, unencapsulated (and therefore avirulent) pneumococci, together with heat-killed encapsulated pneumococci, led to the emergence of encapsulated virulent bacteria, a process he called transformation. This observation remained unexplained until the 1940s, when Avery and coworkers provided conclusive evidence that these mutants had recovered the capacity to produce a capsule by taking up nucleic acids from killed, virulent organisms. Therefore this discovery was the first to identify a functional transferable material (DNA) that encoded for phenotype.
Finally, pneumococcal infections were among the first to be treated with an antimicrobial agent, in this case optochin (ethylhydrocupreine), a quinine derivative. The organism was also among the first to develop resistance to such therapy, resulting in failure of treatment both in experimental animals and in humans.
Microbiology
S. pneumoniae is a gram-positive coccus that replicates in chains in liquid medium but appears as lancet-shaped diplococci in clinical specimens. The organism is catalase negative, but generates hydrogen peroxide (H 2 O 2 ) via a flavoenzyme system and therefore grows better in the presence of a source of catalase, such as red blood cells. Pneumococci produce pneumolysin (formerly called α-hemolysin), which breaks down hemoglobin into a green pigment that surrounds the colonies during growth on blood and chocolate agar plates, a phenomenon still termed α-hemolysis. Pneumococci may be identified in the microbiology laboratory by three reactions: (1) α-hemolysis of blood agar, (2) susceptibility to optochin, and (3) solubility of colonies in bile salts (sodium deoxycholate). Some pneumococci are optochin resistant. A related species, Streptococcus pseudopneumoniae, which is associated with exacerbation of chronic obstructive pulmonary disease or pneumonia, is optochin-susceptible during growth in room air at 37°C but optochin-resistant when grown in the presence of increased carbon dioxide. These factors have led to greater reliance on the use of bile solubility and commercial DNA probes for the ribosomal RNA (rRNA) gene for definitive identification.
Peptidoglycan and teichoic acid are the principal constituents of the pneumococcal cell wall ( Fig. 199.1 ). Peptidoglycan consists of long chains of alternating N -acetyl- d -glucosamine and N -acetylmuramic acid, from which extend chains of four to six amino acids called stem peptides. Stem peptides are cross-linked by pentaglycine bridges, which provide substantial strength to the cell wall. Teichoic acid, a carbohydrate polymer that contains phosphorylcholine, is covalently linked to the peptidoglycan on the outer surface of the bacterial wall and protrudes into the capsule. Teichoic acid and tightly adherent fragments of peptidoglycan make up C-polysaccharide, a substance present in all pneumococci but only in a few species of viridans-group streptococci. C-polysaccharide reacts with acute-phase reactants in the blood during inflammation, including C-reactive proteins. Many proteins are expressed on the pneumococcal cell surface. Of particular importance in the pathogenesis of pneumococcal disease are those that bind to choline, including pneumococcal surface proteins A and C (PspA, PspC); surface adhesins choline-binding protein A (CbpA) and choline-binding protein C; and proteins involved in receptivity to DNA acquisition (or natural competence) ( Table 199.1 ).
TABLE 199.1
Role of Pneumococcal Constituents as Virulence Factors a
| PNEUMOCOCCAL CONSTITUENT | MECHANISM | STRENGTH OF EVIDENCE AS A VIRULENCE FACTOR | |
|---|---|---|---|
| Antibody Prevents Disease b | Mutants Lack Virulence | ||
| Capsular polysaccharide | Prevents phagocytosis; activates complement | 4+ | 4+ |
| Cell wall polysaccharide | Stimulates inflammation by strongly activating complement and stimulating release of cytokines | 0 | ND |
| Pneumolysin | Cytotoxic; activates complement, cytokines | 2–3+ | 2–3+ |
| PspA | Inhibits phagocytosis by blocking activation and deposition of complement on bacterial surface | 2+ | 2+ |
| PspC | Inhibits phagocytosis by binding complement factor H | 1–2+ | 1–2+ |
| PsaA | Mediates adherence | 1–2+ | 1–2+ |
| Autolysin | Causes bacterial disintegration, releases components | 1+ | 2+ |
| Neuraminidase | Possibly supports adherence | 0–1+ | 0–1+ |
ND, Not done; Psa, pneumococcal surface adhesin; Psp, pneumococcal surface protein.
a The grading system is subjective and indicates (on a scale of 1+ to 4+) the stringency and importance of the demonstrated effect. For discussion and references, see the text and Weiser et al.
Nearly every clinical isolate of S. pneumoniae contains an external polysaccharide capsule, but unencapsulated isolates have been implicated in outbreaks of conjunctivitis. Capsules (see Fig. 199.1 ) of repeating oligosaccharides are synthesized in the cytoplasm, polymerized, and transported to the bacterial surface by cell membrane transferases. These polysaccharides are covalently bound to peptidoglycan and C-polysaccharide, which explains the difficulty of separating capsular from cell wall polysaccharide in vaccine preparations. Genetic control of this complex set of events has been elucidated for some serotypes; for example, a cassette of 15 genes that function as a single transcriptional unit is responsible for encapsulation in serogroup 19. At least 97 serotypes of S. pneumoniae have been identified on the basis of antigenic differences in their capsular polysaccharides. Among the multiple genes that encode production of individual capsules, some are specific for individual polysaccharides, whereas others are conserved among nearly all pneumococci and even some other streptococci.
Antibodies induced in rabbits immunized with specific capsular types cause agglutination and create a hydrophobic border around the capsule. This latter reaction, called the Quellung reaction, renders the capsule refractile and therefore more readily visible under the microscope. Because serum antibody is the basis for identifying these types of pneumococcus, they are called serotypes. The American system numbers the serotypes sequentially in the order in which they were identified historically. The more widely accepted Danish numbering system distinguishes 46 serogroups, with groups containing antigenically related serotypes. For example, Danish serogroup 19 includes serotypes 19F, 19A, 19B, and 19C (the letter F indicates the first member of the group to be identified, followed by A, B, C, etc.), which in the American system would be serotypes 19, 57, 58, and 59, respectively. The serotypes that most frequently caused human disease were the earliest to be identified and the first to be assigned numbers, which explains why the lower-numbered serotypes are generally more likely to be implicated in human infection. In the 1930s, serotyping was used to direct therapy with capsule-specific horse antisera. Today, serotypes are important to define epidemiologic and public health surveillance, as targets for vaccines, and for understanding pathogenesis, but not for therapy.
Pneumococci express a competence-sensing protein and internalize DNA from other pneumococci or from other bacterial species. This horizontal transfer of genetic information to pneumococci, called transformation, enables pneumococci to acquire new traits. Of note, a pneumococcus of one serotype can acquire DNA that encodes a different capsular polysaccharide, thereby changing its serotype. This exchange of genetic information occurs under experimental conditions as well as in nature. Thus in addition to a highly conserved genetic core, S. pneumoniae supports a large number of noncore genes that provide remarkable diversity of genetic loci between isolates, particularly related to antimicrobial targets and targets of immune recognition (capsular polysaccharides and adhesins).
Epidemiology
The spectrum of pneumococcal infections can range from asymptomatic colonization to mucosal disease (otitis media, sinusitis, pneumonia) to invasive infections (infection of previously sterile sites). Although otitis media may be the most common clinical manifestation, pneumococcal pneumonia has the greatest impact on morbidity and mortality. Indeed, pneumonia is the leading cause of death of children from infection worldwide, accounting for 1 in 5 deaths, and S. pneumoniae is the leading cause of bacterial childhood pneumonia, particularly severe pneumonia. Pneumococcal pneumonia, sepsis, and meningitis cause more deaths in children younger than 5 years than acquired immunodeficiency syndrome [AIDS], malaria, and measles combined, particularly in resource-limited countries. S. pneumoniae is the leading identified cause of bacterial pneumonia in adults in Kenya and in adults in the United States.
Invasive pneumococcal infections are most prominent at the extremes of life ( Fig. 199.2 ). Consistent results in multiple ethnic and geographic groups highlight the tremendous impact of age on the incidence of bacteremia. In the pre–conjugate vaccine era (before 2000), pneumococcal bacteremia occurred at an approximately 10-fold higher rate among children younger than 2 years than among adults in the general population and in all populations studied, independent of ethnicity (e.g., White Mountain Apaches) or underlying disease (sickle cell disease, splenectomy, human immunodeficiency virus [HIV] infection). These results are likely due to the limited ability of children under 2 years of age to generate protective antibodies to capsular polysaccharides. Infants in the first few months of life tend to be relatively spared in association with the passive transfer of capsule-specific mucosal immunoglobulin A (IgA) and innate factors to the upper respiratory tract by breast milk and specific immunoglobulin G (IgG) to serum transplacentally through cord blood, levels of which decline by 6 months of age. Implementation of pneumococcal vaccination at 2, 4, and 6 months of age for infants in the United States (7-valent pneumococcal polysaccharide-protein conjugate vaccine [PCV7], then 13-valent pneumococcal polysaccharide-protein conjugate vaccine [PCV13] in 2010) has reduced invasive pneumococcal disease by over 90% in young children and by half in older adults.
Among children 6 months to 2 years of age, invasive pneumococcal disease is diagnosed primarily when blood cultures are obtained to evaluate for fever. Many of the affected children with “primary bacteremia” have no apparent focus of infection and are not hospitalized, and one-third resolve spontaneously. Hospitalization is more common with associated underlying cardiac, respiratory, and neurologic disease. Unlike in adults, among whom bacteremia is most often a complication of pneumonia (>80%), pneumonia in young children accounts for 28% to 77% of pneumococcal bacteremias in developing countries and 13% to 60% in more industrialized countries. Primary bacteremia accounts for 61% to 70% of invasive pneumococcal disease in infants in the United States but is uncommonly diagnosed in developing countries. Whether the relatively low levels of serum antibodies to capsular polysaccharides or other protein antigens, or innate factors, in healthy adults underlie their relatively low incidence of pneumococcal disease is not well understood.
Adults older than 65 years comprise about 15% of the population but experience one-third of all cases of invasive pneumococcal disease (approximately 15,000 episodes/yr, with ≥15% mortality). Hospitalization for pneumonia increases from 1.5% to 3.9% per year from age 65 to over 85, particularly among those with diabetes and organ dysfunction. Most invasive cases result from complications of pneumonia (70% to >80%), but 5 to 10 times as many older adults experience pneumococcal pneumonia without bacteremia. In a recent prospective observational study, 8.8% of confirmed cases among adults over 65 years of age hospitalized with serotype-defined pneumococcal pneumonia were complicated by bacteremia, as were 14% to 30% of such cases in a large prospective vaccine trial. Thus the overall mortality associated with S. pneumoniae is likely much greater in the population than the numbers for invasive disease predict. Mortality also increases substantially with age (see Fig. 199.2 ) and is more than two- to fivefold greater among adults with underlying disease (advanced lung, heart, kidney, or liver disease; diabetes; asplenia; solid and hematologic malignancies; immunosuppression) than in healthier older adults.
Deaths from pneumococcal bacteremia tend to occur quickly, often in the first day to week of hospitalization. Despite advances in antimicrobial therapy, the 5% to 10% early mortality with pneumococcal bacteremia has remained constant over the last century ( Fig. 199.3 ). Although chronologic age itself is a factor, most pneumococcal disease and mortality in older adults occur in subjects with diabetes, underlying organ dysfunction (e.g., liver, kidney, heart, lung), humoral immune defects (hypogammaglobulinemia, chronic lymphocytic leukemia, multiple myeloma), and use of immunosuppressive medication ( Fig. 199.4 ). These data suggest an independent or additive contribution of underlying disease to the increased risk of serious infection with age.
Pneumococcal disease is not generally highly contagious. Pneumococci are transmitted from one person to another as a result of close contact, such as among toddlers in daycare centers, but many steps intervene between spread of organisms, colonization, and development of disease. However, community-wide epidemics are well recognized in the meningitis belt in West Africa, where capsular serotype 1 predominates, and smaller outbreaks of disease, often caused by capsular serotypes 1, 5, and 12F, occur among adults in crowded living conditions, such as in military camps, prisons, shelters for the homeless, and nursing homes. In contrast, close contact in schools or the workplace is generally not associated with outbreaks of pneumococcal disease. A very high incidence of invasive infections occurs at all ages among African Americans, Native Americans, and Alaska Natives and Australian Aboriginals, presumably related to both socioeconomic genetic factors. Infants from these populations and other disadvantaged members of developed societies are more likely to be colonized with high numbers of pneumococci, even within the first few weeks of life, adding exposure to their risk for disease.
Pneumococcal colonization and disease occur with a seasonal pattern, with a midwinter increase, although pneumococci can be recovered from healthy children and adults throughout the year. The incidence of pneumococcal otitis media clusters largely from November to April. Bacteremia shows a clear midwinter peak in temperate climates ( Fig. 199.5 ). The association may relate in part to the colder temperatures and lower humidity, which appear to predispose to transmission of respiratory viruses such as influenza, and perhaps closer contact indoors. However, in Houston, Texas, invasive disease in children coincides with the school year, from September through May, sparing the summer months and with no clear midwinter peak. The seasonal distribution is less prominent in tropical climates.
Pathogenetic Mechanisms
Colonization
The prevalence of pneumococcal colonization attests to the success of this organism in adapting to adherence and survival in the nasopharynx. The vast majority of colonizing episodes remain asymptomatic, but most symptomatic infections are likely initiated after asymptomatic colonization. Nasopharyngeal colonization with pneumococci begins in the first weeks of life. The prevalence of colonization increases from less than 10% over the first several months of life to a peak at 70% to 100% at age 1 year, persists through the second and third year of life, and decreases thereafter to adult rates under 5%. Defining such rates depends on the sampling location, methods, and culture, and, more recently, high-throughput sequencing of the 16S rRNA gene and detection of the lytA gene by polymerase chain reaction (PCR). Duration of carriage can range from 1 week to 6 months. In adults, an individual serotype persists for shorter periods, usually 2 to 4 weeks, but sometimes for several months.
Living in resource-limited countries and the presence of siblings in the home are consistent risks for colonization. Crowding (e.g., in home, daycare, barracks), lower socioeconomic status, and ethnic background, as mentioned earlier, have been associated with high carriage rates, as have smoke exposure (passive or active smoking and cooking fires in the home), antibiotic use (with associated risk of carriage of resistant organisms), and respiratory viral infections. Colonization is seasonal, peaking in the winter months but present in children year-round. Colonizing organisms can be transmitted from person to person by an aerosol route from coughing, with particles in the 1- to 5-µm size range depositing primarily in the upper respiratory tract or acquired from saliva and shared drinking vessels.
The frequency of pneumococcal colonization can be influenced by mucosal viral infections and the resident polymicrobial microbiome of co-colonizing organisms, which include other streptococci and Neisseria species, as well as Haemophilus influenzae, Moraxella catarrhalis, Staphylococcus aureus , and other organisms. These bacteria interact and can compete for nutrients and binding sites, produce inhibitory molecules, and modify both local innate and specific immune responses. Bacterial components can have specific interactions with host mucosal constituents. A more diverse nasopharyngeal microbiome was associated with an increased likelihood of colonization in subjects experimentally challenged nasopharyngeally with pneumococcus. Pneumococcal neuraminidases A and B may modify neighboring bacteria and cleave sialic acid on host mucins and glycopeptides to facilitate exposure of N -acetylglucosamine receptors on resting epithelial cells that bind pneumococcal surface-associated proteins, such as pneumococcal surface adhesion protein A. Local inflammation induced by pneumolysin, rhinovirus or influenza infections, and conditions that elicit proinflammatory cytokines, such as tumor necrosis factor-α and interleukin-1, facilitate pneumococcal binding, uptake, and likely migration of the bacteria across the epithelium and endothelium. Such inflammation elicits upregulation of epithelial platelet-activating factor receptor that binds bacterial cell wall phosphorylcholine (ChoP), which intercalates through the capsule. Moreover, sialic acid, lacto- N -neotetraose, and polymeric immunoglobulin receptors on activated epithelial cells bind surface-expressed CbpA to advance transcytosis.
Other innate host macrophage receptors, including scavenger receptor A, mannose receptor, and particularly, macrophage receptor with collagenous structure (MARCO), have been proposed in murine models to enhance mucosal clearance of S. pneumoniae. The role of MARCO may be to enhance interactions between the organisms and other innate receptors, such as Toll-like receptor 2 and nucleotide oligomerization domain protein 2, to promote cytokine and chemokine production. In addition to its cytotoxic and complement-modifying activities, pneumolysin, largely conserved among strains, appears to engage Toll-like receptor 4, and, in conjunction with the cell wall, to limit mortality after mucosal challenge. Finally, specific host IgA bound to the pneumococcal capsule, and potentially to other surface proteins, can be cleaved by bacterial IgA1 protease. The bound but cleaved IgA modifies the surface charge and can enhance binding of encapsulated pneumococci that otherwise bind less well to the epithelium. Pneumococci themselves may also respond to local environmental conditions by increasing expression of ChoP and CbpA to enhance adherence to mammalian cells. Lower expression of capsule may expose epithelial-binding cell wall–bound ligands, but higher capsule expression may facilitate bacterial escape from mucin binding and shedding. Thus S. pneumoniae adapt to and interact with their mucosal milieu to effect colonization, but the presence of inflammation may promote the conversion from asymptomatic colonization to local and systemic disease. When symptomatic infections do accompany colonization, they typically occur within a few weeks of acquiring a new strain rather than from chronic carriage, because antibody production generally follows soon after colonization in children or in adults.
Pneumococcal Capsular Serotypes and Progression From Colonization to Disease
The serotypes of colonizing pneumococci can vary by age and geography and are predictive of the incidence, syndrome, and outcome of pneumococcal infections. Only a relatively small number of the more than 92 capsular serotypes typically cause serious disease, particularly among unvaccinated children. Differences among these capsular types account for approximately 60-fold differences in the invasiveness of the organism. Indeed, capsular serotype appears to be a primary determinant of which strains are likely to remain as colonizers (e.g., 3, 6A, 6B, 9N, 19F, and 23F), which are more likely to progress from carriage to cause more invasive disease (e.g., 1, 5, 7F, 8, 14, 18C, 33F, and 38), and, among those causing invasive disease, which are more likely to cause fatal infections. Based on a process called phase variation, organisms that express less capsule in vitro are more likely to colonize, whereas those producing more capsule show greater resistance to phagocytosis. The clinical distinctions among organisms are associated with their interaction with the host. Colonizing strains are proposed to resist nonopsonic phagocytosis but to be cleared by alveolar macrophages and by complement in blood and, perhaps, lung, thereby resulting in a lower incidence of invasive disease. In contrast, invasive serotypes are less common colonizers that, if not cleared by neutrophils, show increased adhesion-mediated binding to and translocation across the epithelium and are more resistant to killing by alveolar macrophages and complement. However, when colonizing strains do cause invasive disease (e.g., serotypes 3, 6A, 6B, 9N, 19F, and 23F), mortality is higher than with infection with the usual invasive strains, perhaps related to decreased host resistance.
Although the serotype distribution varies by age, these differences are likely not sufficient to explain the increases in mortality (1) from infancy to adulthood, (2) from age 65 to older than 80 years, or (3) in the presence of underlying disease among adults older than 65 years. Disease in immunocompromised adults is more often caused by pediatric serotypes. Thus the interaction between microbial factors, particularly expression of capsular serotype, and the integrity of anatomic and immune host defense determine the development of disease and mortality, but serotype appears to be the primary determinant of outcome in young healthy adults.
These morbid outcomes highlight the importance of understanding, potentially preventing, or at least successfully managing pneumococcal colonization and thereby its consequences. In this context, use of the polysaccharide-protein conjugate vaccines over the past 18 years has been associated with decreased rates of colonization in immunized children. This decrement in childhood colonization and disease has been accompanied by a progressive reduction in adult invasive disease with the pediatric vaccine serotypes. Colonization itself may serve as an immunizing event, and antibodies to pneumococcal neuraminidase A, PspA, and capsular polysaccharides have been detected in serum after carriage. These antibodies, as well as antibody-independent CD4 + T-cell–mediated responses, perhaps involving Th17 CD4 + T cells, may also mediate protection against subsequent colonization and the development of disease and provide alternative approaches to immunization.
Immunologic Mechanisms of Defense Against and Susceptibility to S. pneumoniae Infection
After successful initial colonization, a range of immunologic mechanisms in the host contribute to protection against disease, and a panoply of bacterial factors conspires to evade these defenses. Critical to the process of defense against disease are the interplay among innate immune factors, including complement, antibodies to the bacteria, and the activity of phagocytes. Each is required and none is sufficient alone to clear this invasive mucosal pathogen. The development of pneumococcal disease among patients with specific congenital or acquired immune defects reveals the role of these factors in defense against serious pneumococcal infections.
Antibodies are essential for defense against pneumococcal disease. Antibodies to capsular polysaccharides, the primary virulence factor, are implicated in protection against disease. In support of their protective role, (1) anticapsular antibodies appear in the bloodstream 5 to 8 days after the onset of infection, when fever spontaneously resolves in the absence of treatment; (2) administration of immune horse serum with capsule type-specific antibody in the preantibiotic era was moderately effective in treating pneumococcal pneumonia (see Fig. 199.3 ) ; (3) capsular polysaccharide vaccines elicit both specific antibody and protection against invasive infection; and (4) capsule-specific antibodies support dose-dependent uptake and killing of pneumococci in vitro and protection of experimental animals after pneumococcal challenge in vivo. Failure to produce capsule-specific antibodies in young children and compromised adults results in increased rates of invasive disease. Similarly, Bruton (X-linked) agammaglobulinemia in children and common variable immunodeficiency in adults (all deficient in immunoglobulin M [IgM], IgG, and IgA) predispose to serious pneumococcal infection, as may deficiencies of IgG2 subclass with selective IgA deficiency and selective defects in responses to polysaccharide antigens.
Other than indirect evidence of an association between antibody to pneumolysin or PspA, few data in humans support a protective role for antibody to antigens other than capsular polysaccharides. Maternally derived antibodies to a range of pneumococcal proteins were associated with either an increase or a decrease in infant colonization. Epidemiologic evidence suggests that antibody-independent immune mechanisms provide protection in the population. Indeed, in the preantibiotic era, a proportion of patients recovered from pneumococcal pneumonia without producing measurable amounts of anticapsular antibody. Moreover, young children with documented primary bacteremia and only mild symptoms can clear the infection at a time when their ability to make anticapsular antibodies is very limited. Thus antibody-independent mechanisms, including the effects of CD4 + Th17 T cells during colonization, may also contribute to defense against this pathogen.
Antibodies defend against infection by binding through their variable Fab regions to bacterial surface components, including polysaccharide capsules and proteins. Upon binding, the effector constant Fc region of the antibody engages Fc receptors on phagocytic cells, such as neutrophils and macrophages, thereby providing a bridge between the phagocyte and the organism. Supporting antibody-dependent phagocytosis and killing of the organism, complement also binds to the surface of S. pneumoniae and to complement receptors on the phagocyte. Complement activation by each of the three activation pathways (classical, lectin, and alternative) and deposition on the surface of pneumococci (particularly complement protein C3b) occurs by both antibody-dependent and antibody-independent mechanisms and is largely required for effective uptake and killing of the organism. In general, all three elements are required for clearance of pneumococci: complement, phagocytes, and antibodies; no two alone are sufficient.
Capsule-specific antibodies are proposed to derive from selected subsets of B cells. The ability of polysaccharides to bind and cross link surface antibodies may directly activate naïve and IgM memory B cells, initiating an IgM response. Thus pneumococcal polysaccharides are considered “T-independent” antigens. Decreases in these B-cell subsets are associated with advanced HIV disease and limited IgM responses to pneumococcal vaccine. Impaired vaccine responses and a high incidence of invasive pneumococcal disease characterize patients with common variable immunodeficiency. A decreased frequency of response to pneumococcal vaccine among patients with splenectomy may be related to immunosuppression and underlying malignancy. Very limited IgM responses to pneumococcal vaccine are common with HIV infection and in elderly and, particularly, frail adults.
Although considered T-independent antigens, pneumococcal polysaccharides may also engage and depend upon CD4 + T cells to provide help. Activated T cells can engage B cells directly as well as secrete cytokines to enhance these early responses and to support the switch from production of IgM to IgG and IgA. The mechanisms by which polysaccharides stimulate T cells are controversial. The presence of proteins, particularly those linked directly and covalently to the polysaccharide on the organism or with conjugate vaccines, appears to be required to help B cells produce antibodies in sufficient amounts and of sufficient quality to control these infections. Such enhancement may derive from binding of protein-linked polysaccharides to polysaccharide-specific B cells by their surface antibody. Uptake and processing of the linked protein by these B cells provides protein-derived peptides bound to major histocompatibility complex (MHC) class II molecules on the B-cell surface. Activation of CD4 + helper T cells by these MHC-bound peptides on the B cell may then support the development of these polysaccharide-specific B cells to produce antibodies that bind avidly to the capsular polysaccharides. In this scenario, the processed protein stimulates the T cells. Alternatively, the linked protein may serve to anchor the polysaccharide to MHC class II molecules on the B-cell surface, allowing the polysaccharide itself to stimulate the T cell and enhance antibody production by that B cell. The key to understanding the limitations of antibody responses to capsular polysaccharides and to enhancing these responses with vaccines lies in characterizing and exploiting these protein-polysaccharide interactions. Indeed, the introduction of vaccines with proteins covalently conjugated to polysaccharides has revolutionized vaccine development and efficacy against S. pneumoniae and other encapsulated pathogens by generating capsule-specific antibodies to opsonize and kill the organisms.
The spleen is the principal reticuloendothelial organ that clears unopsonized pneumococci from the bloodstream. In humans, highly opsonized particles are removed from the circulation in part by the liver. However, particularly with limited opsonization, the spleen assumes the most prominent role. Presumably, the slow passage of blood through the spleen and prolonged contact time with reticuloendothelial cells in the cords of Billroth and the splenic sinuses allow the relatively less efficient removal of nonopsonized particles through natural immune mechanisms. Overwhelming pneumococcal infection occurs in children and adults from whom the spleen has been removed or in whom it does not function normally, with an incidence 5- to 15-fold greater than that in other adults and greater yet in children. The herald event in an outbreak of pneumococcal pneumonia in a metropolitan prison was the rapid, septic death of two prisoners who had undergone splenectomy. Pneumococcal disease progressed so rapidly in these cases that pneumonia was not initially detectable clinically or even with certainty by chest radiographs before autopsy. The 35- to 100-fold increase in the incidence of pneumococcal bacteremia or meningitis in children with sickle cell disease is probably due to splenic dysfunction, although other factors, such as antibody and complement abnormalities, may also contribute. Complications of these high-grade infections can include adrenal insufficiency with Waterhouse-Friderichsen syndrome and peripheral symmetrical gangrene with necrosis of multiple digits and limbs.
Factors That Predispose to Pneumococcal Infection
S. pneumoniae is a prototypic extracellular bacterial pathogen. Host defenses against infection rely, as noted earlier, on the interaction between antibody, complement, and phagocytic cells, specifically neutrophils. Both primary (or congenital) and secondary clinical conditions and underlying mechanisms may hamper the immunologic capacity of the host and predispose to pneumococcal infection ( Table 199.2 ). Although these risks include defects in anatomy, antibody production, complement, and phagocytes (typically low neutrophil number rather than impaired function), cell-mediated abnormalities in T and natural killer cells do not figure prominently among them. These predisposing conditions do, however, include underlying liver, kidney, heart, and lung dysfunction; diabetes; alcoholism; and malignancies, particularly in older adults, which may invoke a more subtle constellation of predisposing risks.
TABLE 199.2
Conditions That Predispose to an Increased Incidence and/or Severity of Pneumococcal Infection
| ABNORMALITY | PRIMARY | SECONDARY |
|---|---|---|
| Anatomic | Congenital CSF leak | Poor eustachian tube drainage Traumatic CSF leak Cochlear implants COPD Asthma Preceding viral/influenza infection |
| Antibody defects | Congenital agammaglobulinemia Common variable immunodeficiency IgG2 subclass deficiency (± selective low IgA) Selective hyporesponsiveness to polysaccharides Hyper-IgM syndrome Hyper-IgE syndrome |
CLL Multiple myeloma HIV infection |
| Low complement | Classical pathway (low C2, C1, C4) Alternative pathway (low factors I, H, and B) Low C3 (all pathways) MBL deficiency and polymorphisms |
Nephrotic syndrome Complement consumption |
| Neutropenia | Cyclic neutropenia | Drug-induced neutropenia Aplastic anemia |
| Neutrophil dysfunction | Fcγ receptor IIa ( R131 allele) (low avidity for IgG2) Chédiak-Higashi syndrome |
Diabetes |
| Reticuloendothelial cell defects | Congenital asplenia Hyposplenia |
Splenectomy Sickle cell disease Portal hypertension |
| Combined/other | IRAK4 deficiency (decreased cytokines) | Extremes of age HIV/AIDS Sickle cell disease (spleen, antibody, ± complement) Chronic organ dysfunction (lung, liver, kidney, heart) Chronic alcohol use Solid-organ and bone marrow transplantation Lymphoma |
| Environmental | Environmental smoke Smoking (tobacco) Crowding (daycare, homeless shelters, prison, military training) Stress (military training) Cold season |
C, Complement component; CLL, chronic lymphocytic leukemia; COPD, chronic obstructive pulmonary disease; CSF, cerebrospinal fluid; HIV/AIDS, human immunodeficiency virus/acquired immunodeficiency syndrome; Ig, immunoglobulin; IRAK4, interleukin-1 receptor-associated kinase 4; MBL, mannose-binding lectin.
In addition to the potential defects in antibody production considered previously, the heat-labile complement components of humoral immunity are essential for defense. Of the many possible defects in complement, only those factors required to generate C3b for binding to the bacterial surface and inactivated C3b (iC3b) for phagocytosis and killing by phagocytes are associated with pneumococcal infection. Because the pore-like membrane attack complex (MAC; complement component C6, C7, C8, or C9) that lyses gram-negative organisms cannot pierce the thick pneumococcal cell wall, pneumococci are not killed by serum alone, and defects in these components do not predispose to these infections. In contrast, genetic deficiencies in complement component C3, essential for the activity of each of the three complement activation pathways (classical, lectin, and alternative) are associated with recurrent pneumococcal infection. Deficiencies in complement components C1, C4, and most commonly C2, whether congenital or acquired, are expected to increase susceptibility to pneumococcal infection, although they do so only rarely. The absence of mannose-binding protein in serum, which triggers the lectin complement pathway, may be associated with susceptibility to pneumococcal bacteremia. However, in addition to host defects in complement production, S. pneumoniae expresses a range of proteins that promote degradation of C3 (PhpA) and interfere with deposition of opsonins (PspA, pneumolysin), with activation and generation of convertase (pneumolysin, factor H–binding inhibitor of complement [Hic], and PspC) and with ligands for complement receptors on phagocytes (Hic, PspA/PspC, CbpA, and pneumolysin). These data highlight the importance of the complement system in defense of the host and survival of the bacteria.
Neutropenia of whatever cause is associated with S. pneumoniae infection, although functional leukocyte defects, such as leukocyte adhesion deficiency syndrome (MAC-1 deficiency) generally are not, due to the redundancy of killing pathways in neutrophils. Similarly, defective bacterial killing by polymorphonuclear leukocytes (PMNs), as seen in chronic granulomatous disease, does not predispose to infection with S. pneumoniae ; the absence of catalase renders this organism susceptible to the interaction between its endogenous H 2 O 2 and myeloperoxidase and the halide present in PMNs. At the time of initial hospitalization for acute leukemia, patients are more likely to have infection caused by other gram-positive pathogenic bacteria, akin to the pretreatment situation in multiple myeloma. Limitations in antibody binding to Fc receptors, particularly among subjects with homozygous expression of the R131 allele of the neutrophil Fcγ receptor II that binds the Fc region of IgG2 only poorly, can predispose to these infections.
The susceptibility of elderly persons to pneumococcal pneumonia is multifactorial; the bacteria can exploit the defects described earlier as well as those accompanying impaired functional status, such as weakening of the gag and cough reflexes, malnutrition, and organ dysfunction. The effect of alcoholism is also multifactorial and involves lifestyle (such as cold exposure and malnutrition), suppression of the gag reflex, and possibly deleterious effects on PMN function, although in most instances these alterations have been difficult to attribute to the effect of alcohol alone and may well involve liver disease as well. Heffron cited studies from the preantibiotic era showing a 30% to 50% incidence of alcohol abuse in patients with pneumococcal pneumonia, results confirmed in more recent studies (about one-third of such patients). A disproportionately high number of patients with pneumococcal infection have diabetes mellitus, a condition in which PMN chemotaxis is reduced and phagocytic function is defective, as well as dysfunction of other organ systems and underlying hematologic malignancies that often accompany aging (see Fig. 199.4 ).
Regarding antibody production in older adults, both the differentiation potential of hematopoietic stem cells and the numbers of naïve T cells and B cells decrease with advancing age, as may the subsets of IgM memory B cells proposed to respond preferentially to polysaccharide antigens. These perturbations and decreased responses to capsular polysaccharides have been most closely related to weight loss and frailty rather than to age alone. Studies have been begun to distinguish between innate age-dependent “immunosenescence” and the effects of accumulated underlying disease and frailty on age-related immune dysfunction, particularly decreased responses to vaccination.
As mentioned earlier, prior respiratory viral infection, especially that caused by influenza virus, plays a prominent role in predisposing to pneumococcal infection. Upregulation of surface receptors during viral infection enhances pneumococcal adherence and invasion. Bacteria show greater epithelial adherence and impaired clearance from the airways because of virus-induced damage. Pneumococcal disease is greatly increased in people with altered pulmonary clearance, such as those who have chronic bronchitis, asthma, or chronic obstructive pulmonary disease, and cigarette smoking, including passive exposure, particularly among otherwise noncompromised nonelderly adults. In the United States, socioeconomic factors may contribute substantially to the increased risk for these infections among persons of African-American descent, whereas the very high incidence among certain Native American populations, such as the Navajo, likely reflects genetic and environmental factors.
HIV Infection
Pneumococcal pneumonia occurs 10 to 25 times more commonly with untreated HIV infection than in the general age-matched population (approximately 90 cases/1000 person-years). Rates of invasive pneumococcal disease (e.g., bacteremia) were increased by approximately 80- to 100-fold with advanced untreated HIV disease (overall 176 vs. 3.8/100,000 patient-years for HIV vs. general population, respectively), with recurrence in approximately 20% of cases. HIV-associated risk factors for pneumococcal disease include stage of HIV disease, CD4 + T-cell number, drug and alcohol use, smoking, organ dysfunction, previous pneumonia, and African-American ethnicity, often with impaired antibody responses. Despite a greater than 90% decline in other opportunistic infections, effective antiretroviral treatment has limited the incidence of bacterial pneumonia and pneumococcal disease by only half, or even shown no benefit.
HIV infection is complicated by impaired IgG antibody responses, especially primary responses, to immunization with pneumococcal and other vaccines. That preimmunization levels may be lower, for instance, with influenza or measles suggests a paradoxical loss of specific antibody despite hypergammaglobulinemia. Indeed, selective antigen-specific deficits after immunization were proposed to underlie the increased incidence of pneumonia and, relatively, pneumococcal infection after administration of 23-valent polysaccharide vaccine among HIV-infected adults in Uganda. Pneumococcal vaccine failures were associated with low vaccine responses and opsonic activity. As discussed later, current pneumococcal vaccines show limited or transient protection against invasive disease, highlighting the need to limit HIV-associated immunodeficiency and circumvent related defects with more effective vaccine platforms.
Update: Pneumococcal Gene Mutation Associated with Meningitis
Clinical Syndromes
S. pneumoniae causes infection of the middle ear, sinuses, trachea, bronchi, and lungs by direct spread of organisms from a nasopharyngeal site of colonization and causes empyema by direct extension to the pleural space from the lungs. Infections of the heart valves, bones, and joints are initiated by hematogenous spread, and those of the central nervous system (CNS) and peritoneal cavity by either route. The spectrum of invasive disease includes primary bacteremia without an apparent source or focus of infection. In a population-based study of pneumococcal bacteremia in Israeli adults, 71% of patients had recognizable pneumonia by plain chest radiograph, 8% had meningitis, and 4% had otitis media or sinusitis; 18% had primary bacteremia without an obvious source, a finding that is more common in children than adults.
Otitis Media
Diagnosis of otitis media depends on visualizing the tympanic membrane and testing its function. Virtually every study of culture-proven acute otitis media has shown S. pneumoniae to be the most common isolate or second only to nontypeable H. influenzae , although nearly all of these studies were carried out in populations that had not received conjugate polysaccharide vaccine (see below). Among children ages 6 months to 4 years before widespread use of pneumococcal conjugate vaccine, S. pneumoniae was implicated in 40% to 50% of cases in which an etiologic agent was isolated or in 30% to 40% of all cases. PCR of middle ear fluid from children with otitis media suggests that 50% of specimens that used to be regarded as sterile may be caused by S. pneumoniae or H. influenzae. Pneumococci have historically been the most prevalent pathogens in otitis media in adults as well. Prior infection by a respiratory virus may play a major contributory role by causing congestion of the opening to the eustachian tube, which prevents expulsion of bacteria by normal clearance mechanisms. Prospective longitudinal studies have shown that, when otitis media occurs, it follows within a few weeks after colonization by a new pneumococcal serotype and before the appearance of circulating anticapsular antibody.
Sinusitis
Acute bacterial rhinosinusitis has the same pathogenesis as, and is initially caused by, the same organisms as acute otitis media, with S. pneumoniae and H. influenzae predominating. Obstruction of orifices by viral infection, atmospheric pollutants, or allergens, together with accumulation of fluid in the paranasal sinus cavities, even during simple colds, provides a medium for bacterial proliferation and subsequent acute sinus infection. Acute bacterial sinusitis is most consistently differentiated from viral causes by the persistence of symptoms (≥10 days), severity of symptoms and signs (temperature ≥39°C, purulent discharge or pain for greater than 3 days), and worsening of symptoms (fever, headache, increased nasal discharge) after initial improvement following a simple upper respiratory tract infection. Evolution from acute to chronic sinusitis is associated with more complex bacteriologic findings.
Meningitis
Except during an outbreak of meningococcal infection, S. pneumoniae is the most common cause of bacterial meningitis in adults. In countries that have implemented effective vaccination programs for H. influenzae type b, pneumococci have become the most common sporadic cause of meningitis in children older than 6 months, as well. Even after the widespread use of pneumococcal conjugate vaccine in children, S. pneumoniae remains the most common cause of bacterial meningitis in the United States, but the burden of disease has shifted from young children to older adults.
Meningitis may result from hematogenous spread or by direct extension of bacteria from the sinuses or the middle ear. Favoring the role of direct extension are the association between acute otitis media or sinusitis and infection of the CNS and the well-documented role of S. pneumoniae as the most common cause of recurrent bacterial meningitis associated with head trauma, cerebrospinal fluid (CSF) leak, cochlear implants, or any other break in the integrity of the dura. Favoring hematogenous spread is the association between pneumococcal pneumonia or bacteremia without a known focus and meningitis. In addition, an autopsy study of the temporal bones of children who died of bacterial meningitis showed no evidence for extension from the middle ear, supporting the possibility that even after otitis media, meningitis may develop as a result of bacteremia. Although hematogenous spread to the choroid plexus was originally thought to be responsible for most cases of pneumococcal meningitis, it is now proposed that infection upregulates platelet-activating factor on vascular endothelial surfaces in the meninges and that pneumococci adhere and are internalized by this mechanism. Direct spread to the CNS via lymphatics may also be responsible. Communication through the cochlear aqueduct between the inner ear and the subarachnoid space may explain deafness, a common complication in patients with hematogenous bacterial meningitis.
Once pneumococci appear in the meninges or subarachnoid space, the capacities to escape phagocytosis and produce inflammation are central to the disease process. Intracisternal injection of pneumococcal cell wall constituents in rabbits (principally peptidoglycan, also teichoic acid) causes the CSF abnormalities of bacterial meningitis, presumably through a variety of inflammatory mediators (e.g., complement protein C5a, tumor necrosis factor, interleukin-1, interleukin-6). Interactions with Toll-like receptors 2 and 4 may initiate some protective response, but also stimulate further inflammation.
No distinctive clinical features distinguish meningitis due to S. pneumoniae from that due to other bacteria. Examination of a Gram stain from a centrifuged specimen of CSF provides the correct diagnosis in a large majority of cases with confirmation by appropriate culture. However, bacterial numbers and diagnostic sensitivity are substantially decreased if antibiotics have been given more than 4 hours in advance of CSF sampling. Immunologic detection of pneumococcal capsular material (“bacterial antigen”) generally does not add information beyond what is determined by Gram staining ; the value of available single and multiplex PCR methods is currently under study.
Acute Exacerbation of Chronic Bronchitis
S. pneumoniae is the third most common bacterial cause of exacerbation in patients who have chronic bronchitis, following H. influenzae and M. catarrhalis. Exacerbations are highly associated with acquisition of a new pneumococcal strain.
Pneumonia
You're Reading a Preview
Become a Clinical Tree membership for Full access and enjoy Unlimited articles
If you are a member. Log in here