Abstract and Introduction
Abstract
Streptococcus pneumoniae is the most common cause of community-acquired pneumonia, meningitis, and bacteremia in children and adults. Invasive pneumococcal disease (IPD) primarily affects young children, older adults (> 65 years of age), and individuals with comorbidities or impaired immune systems. Case fatality rates range from 10 to 30% in adults with IPD but are much lower (<>
Introduction
Streptococcus pneumoniae is the most common cause of community-acquired pneumonia (CAP) in adults, accounting for 30 to 70% of cases requiring hospitalization.[1-4] A meta-analysis of 122 reports of CAP between 1966 and 1995 implicated S. pneumoniae in 66% of nearly 7000 cases with an established etiology.[2] In addition, S. pneumoniae is the leading cause of bacteremia,[5-7] meningitis,[8] upper respiratory tract infections, and otitis media[9] worldwide. Bacteremia is present in ∼20% of pneumococcal pneumonias in adults, with case-fatality rates of 10 to 30%.[2,6,10-16] Mortality rates are much lower in children (<>[11,17,18] Invasive pneumococcal disease (IPD), defined as isolation of S. pneumoniae from a normally sterile site [e.g., blood; cerebrospinal fluid (CSF); surgical aspirate; pleural, pericardial, peritoneal, bone, or joint fluid],[19] most frequently affects young children (particularly age 6 to 24 months), older adults (age ≥ 65 years), and immunocompromised individuals (children or adults).[1,20] The World Health Organization (WHO) estimates that 1.6 million people, including up to 1 million children <>[21] with developing countries bearing the greatest burden.[22] In North America and Europe, the annual incidence of pneumococcal bacteremia is 15 to 40/100,000 individuals.[1,7,11,17] In the United States in 2003, 35,000 cases of IPD in adults ≥ age 18 led to 5600 deaths; 44% of cases and 60% of deaths were in adults ≥ 65 years of age (www.cdc.gov/abcs). The incidence is age dependent ( Table 1 ). In the United States, the annual incidence of IPD in children <> 65 years of age.[11,17,23] Regional differences in the incidence of IPD have been noted.[1,7,17,24] In Europe, the annual incidence of IPD in children <>[24-27] In the early to mid-1990s the reported incidence of pneumococcal bacteremia increased in several countries, including the United States,[11,28] Belgium,[25] Sweden,[29,30] Norway,[31] and Denmark.[32] Following the introduction of pediatric heptavalent pneumococcal conjugate vaccine (PCV7) in the United States in February 2000, the incidence of IPD declined substantially in both children and nonvaccinated adults (by herd immunity).[7,15,23,33-36] Unfortunately, the incidence of IPD due to non-PCV7 serotypes is increasing globally.[37,38] Further, fluctuations in incidence of IPD within countries may occur due to clonal spread or other factors (in the absence of vaccine effect).[27,30]
Clinical manifestations of pneumococcal infections are varied and include asymptomatic colonization,[20,39] upper respiratory tract infections, otitis media,[9,40] sinusitis,[41] conjunctivitis,[42,43] bacteremia (with or without a definite site of infection),[5,6] pneumonia,[1,6,10,11] empyema,[44-47] meningitis,[8,48] endocarditis,[41,49-51] septic arthritis,[52] cellulitis,[53] and so forth. In young children, bacteremia without identifiable source accounts for 50 to 70% of episodes of IPD, pneumonia (15 to 25%), meningitis (4%), otitis media or miscellaneous sites (10%).[54,55] In adults, pneumonia accounts for 50 to 80% of episodes of IPD. Other sites of infections include bacteremia without identifiable focus (15 to 20%), meningitis (4 to 8%), miscellaneous sites (2 to 5%).[17,54-57]
Ecology of Streptococcus Pneumoniae Infections
The nasopharynx is the major ecological reservoir of S. pneumoniae; spread from the nasopharynx to lower respiratory tract or other sites may cause invasive disease.[58] Children are the major carriers.[39,59-62] Thirty to 50% of young children (<>S. pneumoniae in the nasopharynx,[39,61,63] compared with carriage rates of only 4 to 12% in adults[39,62,64] and 8.2% in adolescents.[65] Higher rates of carriage (13 to 34%) were noted in adults in select populations.[66-68] Young age (<>[39,59,60,63] Risk factors for NP carriage in adolescents or adults include acute upper respiratory tract infection,[39,65] exposure to passive cigarette smoke,[64,65] and asthma.[65] Transmission from children to siblings, household contacts, or adults is the major cause of IPD.[40,69,70] Interestingly, in Utah, children with IPD due to nonvaccine serotypes tended to be from larger households.[71] Most children have at least one pneumococcal infection (typically of the middle ear) within the first 5 years of life.[72-75] In subsequent years, pneumococcal infections are less common, due to acquisition of humoral immunity.[72] Pneumococcal infections are more common in immunocompromised individuals (children or adults),[76,77] older adults (age ≥ 65 years),[78,79] or in the presence of comorbidities.[55,79]
Pathogenesis of Streptococcus Pneumoniae Infections
NP carriage of S. pneumoniae is required for transmission of bacteria and for invasive disease.[80] Pneumococci bind to mucosal epithelial cells of the nasopharynx.[72] In normal healthy children, NP carriage of pneumococci is transient and is not associated with disease.[21] However, disease is caused by contiguous spread to the sinuses or middle ear, aspiration into the lung, or invasion of the bloodstream.[72] Progression to pneumonia requires additional factors (e.g., antecedent viral infections, lung injury, impaired host defenses, etc.). Clearance of pneumococci is facilitated by both humoral and cellular immune responses involving monocyte/macrophages, polymorphonuclear leukocytes (PMNs), anticapsular antibodies, and lymphocytes.[72] Further, nonimmune factors (e.g., anatomical barriers, cilia, mucins, colectins, surfactant, etc.) are also critical to clear bacteria.[81] Prognosis of pneumococcal infections depends upon both host- and organism-dependent factors.[72]
The polysaccharide capsule serves as a major pathogenic factor for invasive disease by preventing phagocytosis.[72] Humoral antibodies directed against the polysaccharide capsule usually develop within the first 2 years of life; colonization with specific serotypes may elicit serotype-specific humoral antibodies.[72,82] Protection is serotype specific but some cross-serotype protection is found in some cases.[80] However, these anticapsular antibodies (whether acquired naturally or by vaccination) provide incomplete protection against IPD.[80,83,84] Antibodies to serotype 19F reduced colonization rates in some[80] but not all studies.[82] Additional serotype-independent factors are important in preventing or resolving pneumococcal disease and carriage.[80] Components of the pneumococcal cell walls recruit PMNs to the lung, enhance permeability of alveolar epithelial cells, and stimulate cytokine release.[72] The host's primary cellular immune response against S. pneumoniae is mediated by alveolar macrophages (AMs); neutrophils represent a second line of defense.[85] The immune response against S. pneumoniae is complex, involving proinflammatory cytokines released by AMs [e.g., tumor necrosis factor-α (TNF-α) and interleukin-1β (IL-1β)],[81,86] macrophage inflammatory protein-2 (MIP-2),[81] upregulation of myriad cytokines and chemokines (e.g., IL-6, IL-8, and IL-18)[81,87]], and adhesion molecules on endothelial cells.[81] Toll-like receptors (TLRs), expressed on both immune and nonimmune cells, are important to recognize S. pneumoniae and promote bactericidal response by mononuclear cells.[88] Further, granulocyte-colony stimulating factor (G-CSF) recruits and stimulates PMNs, facilitating phagocytosis and oxidative burst.[81] T lymphocyte cells also play a role in eradicating pathogens from the alveolar spaces.[81] Dendritic cells present the antigen to T cells, expanding CD4+ T cell responses, specifically T-helper 1 (Th1) and Th2 phenotypes.[81] Th1 immunity, characterized by production of IL-2, IL-12, IL-18, granulocyte monocyte-colony stimulating factor (GM-CSF), and interferon-γ (INF-γ), is critical to the eradication of pneumococci.[81] Interestingly, deficiency of IL-12 in humans has been associated with recurrent pneumococcal pneumonia.[89] Th2 cells release cytokines that stimulate B cell antibody production, thereby facilitating humoral responses.[81] Other cells/products critical to eradicating pneumococci include antibody- and complement-mediated opsonization; IL-1 receptor-associated kinase-4- and nuclear factor kappaB[90]; and memory T cells (generated in the spleen).[91] In summary, eradication of pneumococci is achieved by myriad interactions involving anatomical boundaries, diverse cells (immune and nonimmune), cytokines, chemokines, and humoral antibodies, and other factors that work in concert. Deficiency of specific immune components (e.g., asplenia, hypogammaglobulinemia, B cell dysfunction, etc.) can lead to recurrent or fatal pneumococcal infections. Further, prognosis of pneumococcal infections depends upon both host- and organism-dependent factors.[13,57,92,93] Antiinflammatory cytokines (e.g., IL-4 and IL-10) have a role in harnessing (blunting) immune responses to infection and can be beneficial or detrimental depending upon the extent of infectious burden and host inflammatory response. Unregulated release of pneumococcal cell wall components during lysis can stimulate brisk inflammatory cytokine responses that may enhance pathological damage and heighten mortality.[81] Optimal response to pneumococcal infections requires a carefully orchestrated and regulated response sufficient to kill the organism without causing excessive injury to the host.
Molecular Epidemiology: Importance of Serotypes
Virtually all strains of S. pneumoniae have a polysaccharide capsule, which is the basis for serotyping.[80,94] Currently, 91 distinct capsular types have been identified.[80] More than 99% of IPDs are caused by S. pneumoniae-containing capsules.[42] Nonencapsulated strains are rare but were implicated in outbreaks of conjunctivitis in military trainees[43] and college students.[42] Globally, ∼20 serotypes account for > 80% of IPD in all age groups; 13 serotypes have been implicated in > 75% of IPD in children.[21] The dominant serotypes associated with IPD worldwide include 14, 4, 1, 6A, 6B, 3, 8, 7F, 23F, 18C, 19F, and 9V.[1,92,95] In young children, the number of serotypes is more limited, with types 6, 14, 18, 19, and 23F predominating.[1] The reasons for the dominance of these serotypes in children, with their immature immune systems, is that these serotypes are less immunogenic than other types.[1,78,96] In most series, more than 80% of IPDs were due to serotypes encompassed in the 23-valent PPV (pneumococcal polysaccharide vaccine), and 40 to 70% were due to serotypes included in the PCV7 (i.e., 4, 6B, 9V, 14, 18C, 19F, 23F).[1,15,54,80,92,96] A major shift in the distribution of serotypes has been noted over time. In the beginning of the twentieth century, serotypes 1, 2, 3, and 5 accounted for up to 75% of bacteremic cases in the United States and Europe.[1,97] Today, types 2 and 5 are rarely isolated in Western countries, and type 1 is uncommon.[1] Instead, other serotypes have increased in prevalence. Interestingly, types 1 and 5 are common today in developing countries.[1] Substantial differences in the distribution of serotypes have been noted in different geographical regions; further, the distribution may change over time.[1,29] Serotype 3 has decreased in frequency over the past few decades, whereas type 14 has become more prevalent.[1,29,31,32,96,98] In Sweden, the incidence of type 14 isolates increased threefold from 1987 to 1992; type 1 increased 10-fold from 1992 to 1997.[30] The increase in type 14 reflected a hypervirulent clone that has also been found in the United States and Canada.[30] Clonal expansion of other serotypes (e.g., type 1,[99] serotypes 15 and 33[38] and type 19A[74,100-103]) has been noted in the United States. Changes in serotype distribution can be modified by antibiotic usage and vaccination patterns. In recent years (following the introduction of PCV7), nonvaccine serotypes are increasing in frequency.[37,103,104] In a review of pneumococcal pneumonia in children in Utah, nonvaccine serotypes represented 49% of isolates from 1997 to 2000 and 88% of isolates from 2001 to 2006.[37] Different serotypes exhibit differences in attack rate and colonization,[92,95,105-107] case-fatality rates, and clinical expression of disease.[15,56,57,92,105] In a study of pneumococcal pneumonia in children, serotype 3 was 15 times more likely to cause necrotizing pneumonia compared with other serotypes.[37] Further, serotype 3 was associated with an increased risk of empyema and need for surgical procedures compared with other serotypes.[37] Other investigators cited higher case fatality rates with IPD caused by serotype 3,[13,57,92,93] although this was not a consistent finding.[15,108] Sandgren et al evaluated 273 invasive (257 from adults) and 246 NP isolates (all from children) of S. pneumoniae in Stockholm in 1997.[95] The isolates formed two major classes: one class comprised serotypes 1, 4, 7, and 9V and was highly clonally related; a second class caused invasive disease but also was common in carriage (including type 6A, 6B, 14, and 19F isolates) and was genetically more diverse.[95] Serotypes 9V accounted for 7% of invasive isolates but only 2% of carrier isolates. By contrast, serotypes 6A and 19F accounted for 34% of isolates among carriers but only 3% of invasive isolates. Specific clones within the same serotype exhibited different abilities to cause invasive disease. Further, isolates belonging to the same clone that exhibited capsular switch displayed the same disease potential. In a subsequent study by these investigators, serotypes with the highest invasive potential (e.g., types 1 and 7F) were associated with low fatality rates, whereas serotypes 3, 6A, 6B, 11A, 19F, and 23F were associated with low invasive potential but caused more severe disease and increased mortality.[56] Isolates with high invasive potential behave as primary pathogens, whereas strains with low invasive potential behave as opportunists (e.g., among patients with underlying disease).[56] An international study from five countries from 1993 to 1995 found that serotype 14 was most common, but type 3 dominated in the fatal cases and in the United States and Spain, countries with the highest mortality rates.[92] A study of 494 adults with IPD noted correlations between age and serotypes causing IPD.[56] Serotype 1 was associated with IPD only among patients <> 65 years old (78%).[56] In one study, serotype 12 was associated with lower mortality,[57] whereas a prospective international study found no association with serotype and mortality when other risk factors were taken into account.[15] Nonetheless, an association between serotype and enhanced virulence and severity of disease is plausible. In a murine model, cytokine proinflammatory responses were more robust in clones with less invasive potential,[109] suggesting that differences in innate immune responses to specific clones (strains) may explain differences in the clinical expression/severity of disease. Serotype prevalence varies among geographic regions and may change over time in response to selective pressure or clonal spread.[30,95,110-113] Further, capsular switching may occur, thus providing a survival advantage to the organism (by eluding capsular-specific opsonizing antibodies), and may facilitate survival of specific clones.[56,95] In addition to capsular types, other properties of the organism (e.g., virulence factors, etc.)[30,114] may influence disease severity and invasive potential. However, host factors are likely more important than serotype regarding severity of disease and mortality rates.[15]
Risk Factors for Invasive Pneumococcal Disease
The incidence of IPD is much higher at extremes of age (<>[17,57,73-75,115,116] (Table 2), in patients with comorbidities or defects in immune defenses[1,17,57,71,76,115] (Table 3 and Table 4), and in the winter months.[1,117] In both children and adults, chronic illness is the strongest risk factor for IPD[71,76] (Table 3). Invasive pneumococcal infections usually reflect transmission from NP carriage in children. Outbreaks of IPD may occur in crowded, closed settings such as schools,[42] day care centers,[115,118,119] households with multiple children,[40,75,115,120,121] long-term care facilities (LTCFs),[68,122-124] closed communities,[125] military camps,[43,126] shelters,[127] jails,[128] and hospitals.[119,129,130] Interestingly, South African women infected with human immunodeficiency virus (HIV)were more likely to develop IPD with pediatric serotypes compared with HIV-infected men, likely because of women's closer proximity to small children.[131] Widespread pneumococcal vaccination may reduce risk of IPD in closed populations[68,132] and the community at large.[23,55]
Infants <>[71,115,133,134] However, the risk of IPD is increased among preterm and low birth weight infants <>[115] In one study, day care attendance was associated with increased risk of IPD in infants between 6 and 23 months old, but not in older children (age 2 to 5 years).[115] These trends reflect natural acquisition of immunity (usually by 2 to 3 years of age), that is durable. A history of tympanostomy tube was associated with an increased risk.[71] In adults, advanced age predisposes to IPD, owing to the presence of comorbidities or immune senescence.[17,76,124]
Ethnic and Socioeconomic Factors
Regional and racial differences in the incidence of IPD have been noted globally.[24,74,76,135] Pneumococcal infections are more common in indigenous peoples of Alaska[136-138] and the Canadian arctic,[138] Inuits in Greenland,[139] American Indians (White Mountain Apache and Navaho),[82,140-142] blacks in the United States,[17,74,76,135] Australian aborigines,[143] Maoris of New Zealand,[144] and Bedouins of Israel.[145] Socioeconomic factors likely are responsible for the higher incidence in these groups, but genetics may play a contributory role. Genetic factors (e.g., polymorphisms) may play a role in susceptibility to IPD,[146-149] but one large study from Denmark found no higher risk of IPD among relatives apart from risk associated with sharing the same household.[75]
Risk Factors in Immunocompetent Individuals
Exposure to cigarette smoke and multiple children in the household are risk factors for IPD in healthy children.[121] In otherwise immunocompetent adults, the incidence of IPD is increased with the following comorbidities: alcohol abuse,[76,142,150-152] congestive heart failure,[76,135,142,153] chronic lung disease,[76,151,153] cigarette smoking,[151,153,154] asthma,[155] recent influenza infection,[156] diabetes mellitus,[76,135,157] institutionalization,[124,153] neurological disorders,[153] male gender,[154] and black race[17,76] (Table 2). Among immunocompetent adults 18 to 64 years of age, current cigarette smoking was the strongest risk factor for IPD.[154] Further, a dose-response relation was found between number of cigarettes smoked and risk of IPD.[154] Multiple comorbidities or age ≥ 65 years amplify the risk.[17,57,76,158]
Immunosupressed Individuals
Patients with primary or acquired immune deficiencies have a heightened risk for IPD.[76,90,135] The risk is highest among patients with B cell defects (due to intrinsic B cell anomaly or impaired T cell helper activity) or deficiencies of early components of the classical pathway of complement.[90] Disorders associated with increased risk of IPD include asplenia,[159-161] hemoglobinopathies (particularly sickle cell disease),[151,162-164] hematological[76,135,165] or solid[76,135,166,167] malignancies,[168-170] organ transplant recipients,[168,171-173] HIV infection,[151,174] and primary or acquired immunodeficiency states or receipt of immunosuppressive drugs (including corticosteroids).[90,168-170] Each of these disorders is discussed separately in the sections that follow. Recurrent episodes of IPD are uncommon in immunocompetent hosts but may be observed in patients with severe immunodeficiency states (e.g., HIV infection or asplenia).[17, 54, 175-177].
Sickle Cell Disease
The incidence of IPD in children with sickle cell disease (SCD), particularly those with homozygous disease (SS), is 30- to 600-fold higher than in individuals of comparable age and race without SCD[151,162-164,178,179] (Table 4). Pneumococcal septicemia and meningitis are important causes of death in SCD patients, with case/fatality rates of 15 to 35%.[164,180-186] The risk of IPD is highest in children <>[162,164,181] Disease severity is usually worse in patients with homozygous (SS) and heterozygous (SC) disease. In one series, hypotension was more common with SS, whereas acute chest syndrome and otitis media were more common findings in patients with SC.[164] The incidence of IPD and mortality due to bacterial infections in SCD has declined over the past 2 decades,[164,182,184,187] likely reflecting the impact of penicillin prophylaxis, earlier recognition and treatment of infections,[181,182,186,188] vaccines,[162,181,189,190] and improved medical care.[162]
However, even with prophylactic measures (e.g., antimicrobial prophylaxis and vaccination with the 23-valent polysaccharide pneumococcal vaccine (23PPV) at age ≥ 2 years), the incidence of IPD was 10-fold higher among SCD patients compared with controls.[164] Further, many children with SCD do not receive antibiotic prophylaxis.[163,188] Since 2001, the use of PCV7 has markedly reduced the incidence of IPD in children with SCD.[162,179] From 1995 to 2000, the rate of IPD (per 1000 patient years) in children ages 0 to 10 years with SCD in Atlanta was 1.7; by 2001-02, the rate had declined to 0.5.[179] In Tennessee, the rate of IPD decreased by 93.4% in SCD children ≤ 5 years old.[162] Antibody responses are enhanced by using both PCV7 and 23PPV vaccines in children or young adults with SCD.[191] Optimal duration of antibiotic prophylaxis is uncertain.[163] Because the rate of IPD declines substantially in older children (> 5 years),[164,181,192] it is reasonable to discontinue antibiotic prophylaxis after age 5 provided (1) patients have not had prior severe IPD, (2) vaccination has been administered, and (3) splenic function is adequate.[193,194] Further, penicillin prophylaxis may not be necessary in countries or regions associated with a low rate of pneumococcal infections (e.g., Africa).[195]
Splenectomy or Asplenia
Splenectomized patients or those with functional asplenia are at increased risk for life-threatening infections due to encapsulated bacteria (including pneumococcus).[160,169,170,196,197] The risk of IPD after surgical splenectomy among nonvaccinated children ranges from 1 to 9%.[160,196,198] Most IPD occur within 1 to 2 years of splenectomy,[160] but the risk may persist for > 15 years in some cases.[197] The incidence of IPD is even higher (up to 8.5%) among children with congenital asplenia.[160,161,199] Further, the case-fatality rate of IPD is higher among asplenic children.[160] Administering vaccines and prophylactic antibiotics reduces the risk.[160,198] All asplenic children <>[160] However, response to vaccination may be blunted in asplenic patients.[169] Antibiotic prophylaxis is warranted, but appropriate duration or therapy is controversial.
Risk in Solid and Hematologic Malignancies
The incidence of IPD is increased in patients with solid or hematologic malignancies[76,135,165-170,200] ( Table 4 ). Kumashi et al reported 135 consecutive episodes of S. pneumoniae bacteremia in 122 cancer patients.[166] Fifty-two percent of patients had hematological malignancies; 48% had solid cancers. Only 24 episodes (18%) occurred during neutropenia. Pneumonia was present in 67%; infected catheters accounted for 18% of episodes. Most (88%) episodes of bacteremia were community acquired. Overall, 19 patients (16%) died within 2 weeks of diagnosis. In a study of 56 cancer patients with pneumococcal bacteremia, the incidence was highest (> 1000 cases per 100,000) in the following groups: Hodgkin disease postsplenectomy, multiple myeloma, and chronic lymphocytic leukemia.[167] In a recent study, German investigators cited a 10-fold increased incidence of IPD in children with acute lymphoblastic leukemia compared with the general pediatric population.[200] This increased incidence of IPD reflects chemotherapy-induced immune aberrations, including: loss of B or T cell activity, neutropenia, chemotherapy-induced hyposplenism, impaired antibody responses to vaccines, and disruption of respiratory mucosa ciliary function.[76,166]
Organ Transplant Recipients
The risk of IPD is much higher in recipients of hematopoietic stem cell (HSC)[168,172,173] or solid organ[171,201] transplants[168,172,173] compared with healthy controls (Table 4). Infection risk depends upon intensity of immunosuppression and environmental factors and usually occurs > 3 months posttransplantation.[171,172] The incidence is highest in allogeneic HSC recipients with chronic graft versus host disease (GVHD).[172,173,202-204] Compared with the general population, the relative risk (RR) of IPD was 30.2 among HSC recipients[172] and 12.8 among solid organ transplant (SOT) recipients residing in the same geographic region.[205] Serotypes implicated among transplant recipients are similar to those reported in immunocompetent patients.[171,172,205] Vaccination is important for all transplant recipients (SOT and HSC).[206,207] Efficacy of vaccination may be blunted, however, by depletion of T and B cells among HSC recipients or the effects of immunosuppression on B cells in SOT recipients.[171,172,207,208] The conjugate vaccine (PCV7) has enhanced immunogenicity compared with 23PPV and elicits a T cell and memory response.[171,207] In a randomized trial, 3 doses of PCV7 administered at 3, 6, and 12 months after HSC conferred protection in most patients (72 to 100% for different serotypes) by 12 months after transplantation.[209] Although the optimal vaccination schedule, type of vaccine, and efficacy among transplant recipients have not been validated in clinical trials,[207] sequential doses of 23PPV at 12 and 24 months post-HSC transplant or combinations of both 23PPV and PCV7 are reasonable.[172,207] For small children and patients (all ages) with chronic GVHD, administering 3 doses of PCV7 starting at 6 to 12 months after HSC is reasonable.[207] Given the high risk of IPD among allogeneic HSC recipients and chronic GVHD, long-term antimicrobial prophylaxis may be warranted in this group.[203,206] The choice of antibiotic depends on local antibiotic resistance patterns.[207]
Human Immunodeficiency Virus Infected
In the era prior to highly active antiretroviral therapy (HAART), the rate of IPD among adults with HIV infection or acquired immunodeficiency syndrome (AIDS) in the United States or Europe was > 40 times higher than age-matched populations[19,151,174,210-214] (Table 4). The incidence of IPD was highest among injection drug users with HIV.[210,211] In HIV-infected children, the incidence rates of IPD are exceptionally high (ranging from 183 to 18,500 episodes per 100,000 child years.[215-219] Recurrent episodes of IPD are more common among HIV-infected patients.[54,210,214] A study of Gambian women with HIV/AIDS noted increased rates of NP colonization, often with pediatric serotypes.[220] Surveillance data from the ABC Study in the United States noted that HIV-infected persons accounted for 15 to 20% of cases of IPD from 1998 to 1999.[221]
Since the introduction of HAART, marked declines in the incidence of IPD[210-212,222-224] have been noted in HIV-infected adults in developed countries (Table 4). Fewer data are available in children, but epidemiological studies cited severalfold reductions in the incidence of bacteremias[225] or pneumonias[225,226] in children between the pre-HAART and HAART eras.[227] Unfortunately, in areas of the world with a large burden of HIV infection, the incidence of IPD may be increasing.[228] Some studies reported lower mortality rates for IPD, including meningitis, among HIV-infected patients.[210,229] This lower mortality rate may in part reflect a blunted inflammatory response to S. pneumoniae [230] and younger age of HIV-infected patients.[210]
In the ABC surveillance study from 1998 to 1999, the distribution of serotypes causing IPD differed among adults with HIV/AIDS compared with adults with no underlying disease.[221] The serotype distribution among HIV/AIDS patients was similar to those with hematogenous cancers.[221] Similarly, a series of IPD in South Africa noted differences in serotypes and antimicrobial resistance patterns among HIV-infected and non-HIV-infected persons.[217,228,231]
The use of trimethoprim/sulfamethoxazole (TMP/SMX) prophylaxis may select for more resistant pathogens,[210,232] but prior use of TMP/SMX prophylaxis was not associated with TMP/SMX susceptibility in a large cohort (n = 416) of HIV-infected patients with pneumococcal bacteremia.[221] However, in South African children with IPD, resistance to penicillin, TMP/SMX, and multidrug resistance were more common in HIV-infected children.[217]
Pneumococcal vaccination is recommended for HIV-infected adults and adolescents with CD4+ lymphocyte counts > 200 cells/µL,[233] but data supporting efficacy are limited.[210,223] Immune deficits associated with HIV infection may dampen the antibody response to 23PPV.[230,234-236] The use of 23PPV did not reduce the incidence of IPD, all-cause pneumonia, or mortality in a cohort of HIV-infected adults in Uganda.[237] Retrospective studies in the United States suggested that 23PPV protects against IPD among certain groups of HIV-infected patients, specifically those with CD4 counts ≥ 200 cells/µL[238] or ≥ 500 cells/µL[212] at the time of vaccination or those receiving HAART.[212,223] These studies were not controlled for comorbidities. A recent prospective study in the United States (the Veterans Aging Cohort 5-Site Study) found that vaccination with 23PPV reduced the risk of pneumonia in HIV-infected adults; current smoking, low hemoglobin level, and low CD4 cell count significantly increased the risk.[239] Among non-HIV infected patients, vaccination with 23PPV did not confer protection (possibly because of lack of statistical power).[239] Benefits associated with HAART may reflect its effects on improving B cell function[240] and qualitative and quantitative responses to pneumococcal antigens.[241] One randomized trial in HIV-infected adults found that vaccination with two injections of either PCV7 or 23PPV 2 months apart elicited higher antibody responses compared with placebo/23PPV.[242] Vaccination with 23PPV after previous vaccination with PCV7 enhanced antibody response in HIV-infected adults.[243] In HIV-infected children, PCV7 has been shown to be safe.[215,244,245] PCV7 is immunogenic in HIV-infected children, but less so than in HIV-uninfected children.[215,244-246] In children with HIV-infection, a positive correlation between antibody concentration elicited by PCV7 and duration of HAART was found.[245] However, the functional activity of pneumococcal antibodies elicited by PCV7 was lower in HIV-infected compared with noninfected children.[247] Clinical efficacy of PCV7 in HIV-infected patients remains uncertain. However, a South African trial noted significant reductions in IPD and pneumonia in HIV-infected children with a nine-valent vaccine.[248] This effect was attenuated at 5 years compared with non-HIV-infected children.[249] The impact of PCV7 in communities with high rates of HIV infections is not known. In the United States, the rate of IPD in HIV-infected adults (aged 18 to 64 years) declined by 19% from 1998-99 to 2003; vaccine-type IPD fell by 62% in this group.[73] The optimal vaccination policy for HIV-infected persons has not been elucidated. However, universal vaccination of HIV-infected patients is reasonable to reduce colonization and infection.[215]
Risk Factors for Mortality in Pneumococcal Pneumonia
Mortality for bacteremic pneumococcal pneumonia ranges from 10 to 30% in adults and <>[2,6,10,11,17,250] Case fatality rates for meningitis range from 16 to 37% in adults[8,34] and 1 to 2.6% in children.[34,251] Extrapulmonary manifestations will not be further addressed here. Mortality rates are much higher in the elderly and patients with comorbidities. Although disparate results have been noted, factors associated with higher mortality in bacteremic pneumococcal pneumonia include age (> 65 years),[2,5,11-13,15,252,253] multilobar involvement,[2,12,14,252,254] renal failure,[11,12,17] leukopenia,[14,252,255-257] alcohol abuse,[6,257] immunosuppression,[5] chronic cardiac disease,[11,17] malignancy,[17] chronic pulmonary disease,[2,254] residence in a nursing home,[2] serious underlying disease,[5,14,15] need for intensive care unit (ICU)[6] or mechanical ventilation,[2,253] shock,[14,257] high acute physiology scores,[2] severity of disease,[15] and treatment with parenteral nutrition.[253] Discordant therapy was associated with higher mortality in some[254,258] but not all[5,259,260] studies. The impact of antimicrobial resistance and discordant antibiotic therapy is discussed in depth in the next article. HIV-infected patients have lower mortality in some studies,[14] likely due to younger age. However, when patients were stratified by clinical status, patients with AIDS had far higher mortality than HIV-infected persons without AIDS.[17,174,211,215] Interestingly, in a prospective, international study of IPD in adults, rigors and chest pain were associated with a lower mortality.[15]
Clinical features of pneumococcal pneumonia are reviewed in detail elsewhere.[1,6,10] We will limit our remarks to changes in the clinical presentation of pneumococcal pneumonia within the past few years (concomitant with the use of PCV7). Recent studies cited a marked increase in pneumococcal empyemas[44,47,261-263] and necrotizing pneumonias[37,45,262,264] in children, which reflects replacement by non-PCV7 serotypes (particularly serotypes 1, 3, and 19A).[37,261,265,266]
Laboratory Diagnosis
Nonbacteremic pneumococcal pneumonia may be difficult to diagnose. Gram stains and microbiological cultures are the mainstays of diagnostic tests but are positive in fewer than 50% of cases of pneumococcal pneumonia.[10] S. pneumoniae appear as lancet-shaped gram-positive diplococci. For bacteremic pneumococcal pneumonia in adults, sputum Gram stain and cultures had sensitivities of 80% and 93%, respectively, provided an adequate specimen was produced prior to therapy.[10] In actual clinical practice, sensitivity is lower (<>[10,267,268] Rapid detection of pneumococcal antigens in urine,[269,270] CSF,[271] or pleural fluid[272] may be helpful in selected patients (especially patients with meningitis or those who have received prior antibiotics).[267] Latex agglutination tests are widely used to diagnose pneumococcal meningitis,[271] but the value of these tests for urine or blood is controversial.[267] An immunochromatographic test (NOW Streptococcus pneumoniae Antigen Test, Binax, Inc., Scarborough, ME), detects the C-polysaccharide wall antigen of S. pneumoniae.[272,273] This urinary antigen test enables a rapid diagnosis (within 15 minutes) of pneumococcal pneumonia, but was less sensitive than sputum Gram stains in some studies.[270,273] False positives, rare in adults,[273] may be noted in children with NP carriage of S. pneumoniae.[274,275] The combination of urinary antigen detection and sputum Gram stain increases the sensitivity[270] but is expensive and time consuming in clinical practice. Binax NOW may be most useful for rapid diagnosis of pneumococcal meningitis, empyema, or pneumonia in high-risk patients in whom adequate sputum is unavailable.[270,272,273] In children, Binax NOW lacks specificity and cannot distinguish colonization from infection.[276] None of these techniques replaces culture, the only technique that allows antimicrobial susceptibility testing.
Pneumoccocal Vaccines
Currently, two pneumococcal vaccines are available. The 23-valent PPV (23 PPV), composed of purified free polysaccharides derived from the surface capsule of the bacterium, was introduced in 1983.[277] These polysaccharide antigens elicit a T cell independent immune response and are therefore poor inducers of immunologic memory.[1,245] Recommendations for using 23PPV in the United States are listed in Table 5.[278] Indications for and extent of usage of 23PPV vary considerably among countries.[277] 23PPV was efficacious and cost-effective in reducing the incidence of IPD in adults[1,279,280] and may prevent outbreaks of pneumococcal pneumonia in institutional settings (e.g., nursing homes).[68,132] The vaccine confers 60 to 80% protection against IPD in young healthy adults[1] and elderly adults[1,280-283] but is less effective in immunocompromised patients.[237,280,284] Further, 23PPV is less effective in preventing nonbacteremic pneumonia[282,285] or noninvasive infections (e.g., otitis media, conjunctivitis).[286,287] Studies in elderly adults found that high serum antibody titers persist for 1 to 2 years[288,289] but wane substantially over 5 years,[290] as does the clinical effectiveness of the vaccine.[280] In the United States, vaccination with 23PPV is recommended for all adults ≥ age 65 years and for high-risk individuals 2 to 64 years of age[277] (Table 5). In Europe, indications for vaccination are variable among countries. In one survey, 17 of 21 countries recommended 23PPV for all adults ≥ 65 years of age.[291] Recommendations for the use of PCV7 in Europe vary among countries.[291] Revaccination for at risk adults after 5 years is safe and immunogenic,[292,293] but vaccine responses are attenuated.[294] Self-limited local reactions at the injection site following revaccination are more common (particularly in immunocompetent patients) but are not a contraindication to revaccination.[21,295] Unfortunately, many patients who are candidates for 23PPV remain unvaccinated.[296]
The 23PPV is poorly immunogenic in children ≤ age 2 years old, but PCV7 (introduced in the United States in February 2000) elicits good antibody responses in infants and young children.[21,297] Conjugation of the capsular polysaccharide to a protein carrier elicits T cell responses that establish immunologic priming and a memory response.[21,298] The serotypes incorporated into PCV7 include 4, 6B, 9V, 14, 18C, 19F, and 23F.[299] In 2000 these seven serotypes accounted for > 80% of IPD in children in North America[55,300] and were also prevalent in elderly adults.[301] Currently, PCV7 is the only commercially available pneumococcal conjugate vaccine, but 10- and 13-valent conjugate vaccines may be available soon.[21] As of January 2007, PCV7 was licensed in more than 70 countries.[21] In 2000, recommendations for PCV7 use were published by the American Academy of Pediatrics Committee on Infectious Disease[194] and Centers for Disease Control and Prevention[297] (Table 6).[302] In Europe, recommendations for the use of PCV7 in Europe vary among countries.[291] In 1997, the WHO advised that PCV7 vaccination in children should be a priority, particularly in countries where mortality was high among children <>[21] Additionally, PCV7 should be prioritized in countries with a high prevalence of HIV, sickle cell disease, or other high-risk populations.[21] Although precise schedules of vaccination differ, most countries administer 3 doses in infants within the first 6 months of life.[299,303] In the United States, a booster dose of PCV7 is given at 12 to 15 months, and 23PPV is given at 2 years of age to broaden serotype coverage.[297] Among developing countries, 3 doses of PCV7 to infants were immunogenic but data regarding clinical outcomes are limited.[21] A 9-valent PCV that contains serotypes 1 and 5 in addition to the serotypes in PCV-7 was efficacious in Gambian children[304] and HIV-infected children from South Africa,[248] but it is not commercially available.[300,305] The duration of protection against IPD due to PCV7 serotypes is at least 2 to 3 years, but probably is considerably longer.[21] Although data are limited, PCV7 may have a role in elderly or immunocompromised adults. A recent randomized trial in vaccine naive adults ≥ 70 years of age assessed the impact of initial vaccination with either 23PPV or PCV7, followed by a booster vaccine (PCV7 or 23PPV) administered at 1 year. The seven serotypes encompassed by PCV7 were assessed. Initial vaccination with PCV7, followed by 23PPV at 1 year, elicited antibody responses that were comparable or higher than 23PPV alone. Importantly, initial vaccination with 23PPV followed by PCV7 induced lower antibody responses than PCV7/PCV7 or PCV7/23PPV. Thus 23PPV was ineffective as a priming dose and may induce hyporesponsiveness to subsequent booster doses (likely by depleting polysaccharide-specific memory B cells). Others have shown that revaccination with 23PPV after 5 years elicits lower antibody levels compared with the priming dose.[294]
Following the use of PCV7 in children, marked declines in IPD were observed among all age groups (even in nonvaccinated adults) in North America,[21,23,35,36,55,57,138,299,305-307] including high-risk[189] and immunosuppressed[57,162,308] individuals.[21,191,299,305,307] In the United States, rates of vaccine-type IPD declined by 62% among people ≥ 5 years old between 1998-99 and 2003.[307] In Canada, the rate of IPD in adults over 65 declined by 63% between 1998-2001 and 2004.[309] Importantly, PCV7 reduced both colonization and infection with S. pneumoniae.[21,57,308,310,311] By contrast, the incidence of IPD increased in Spain following the introduction of PCV7 and was associated with emergence of nonvaccine serotypes.[112] In Alaskan native adults, colonization with vaccine-serotypes declined from 28 to 4.5% after introduction of PCV7 in children.[312] Further, after 2001, the incidence of pneumococcal meningitis declined in the United States in both children and adults.[251] The use of PCV7 in the United States led to reduction in racial disparity in IPD.[74] However, in recent years, case-fatality rates have increased, reflecting a higher proportion of cases with comorbidities or non-PCV7 serotypes.[57] In the United States and globally nonvaccine serotypes account for an increasing proportion of IPD, acute otitis media, and NP colonization.[36,38,55,73,101,102,112,137,307,313,314] Ominously, in a study of Alaskan native infants, IPD due to nonvaccine serotypes increased by 140% since 2004 compared with the prevaccine period.[137] During the same period (after 2004), IPD due to PCV7 serotypes declined by 96%.[137] Further, the proportion of IPD cases with empyema increased from 2 to 13%.[137] Several other studies cited an increased incidence of empyema in children in recent years due to nonvaccine serotypes.[37,44,47,261-263] A prospective study of IPD from eight children's hospitals in the United States noted emergence of replacement serotypes 15 and 33; two dominant clones were observed.[38] Currently, 19A has been the most common replacement serotype in the United States,[74,101-103] whereas serotypes 1 and 5 have been most common in Spain.[112] In some areas, this expansion has been clonal.[112,137]
Vaccination strategies in both children and adults will continue to evolve. Serotypes affecting adults and children differ, mandating different strategies for specific populations. Changes in the distribution of serotypes following PCV7 will be a challenge for future vaccination strategies. Conjugate vaccines containing up to 13 serotypes are in development to improve coverage in adults and children on a global basis.[21,315] It should also be emphasized that vaccinating elderly adults against influenza has reduced the risk of all-cause pneumonia,[316,317] and may protect against IPD.[318]
Source : http://www.medscape.com/viewarticle/702635
Pulmonary alveolus — An alveolus (plural: alveoli), is an anatomical structure that has the form of a hollow cavity. In the lung, the pulmonary alveoli are spherical ... > read more
Gas exchange — Gas exchange or respiration takes place at a respiratory surface; a boundary between the external environment and the interior of the body. For ... > read more
