key: cord-0036546-uyddvm0x
authors: Annamalay, Alicia; Le Souëf, Peter
title: Viral-Bacterial Interactions in Childhood Respiratory Tract Infections
date: 2017-05-05
journal: Viral Infections in Children, Volume I
DOI: 10.1007/978-3-319-54033-7_8
sha: cbcd1a019ffc2158d9f6c58e251f8b30d297a6bb
doc_id: 36546
cord_uid: uyddvm0x

Acute respiratory infection (ARI) is an important cause of childhood morbidity and mortality worldwide. ARIs are caused primarily by viruses and bacteria that are often co-detected in respiratory specimens. Although viral-bacterial co-infections are frequently reported in children with ARI, their clinical significance and the mechanisms leading to ARI are not well understood. The respiratory tract is a reservoir of a diverse community of microorganisms, including both commensals and potential pathogens and there is growing evidence that the interactions between viruses and bacteria play a key role in the development of ARI. A better understanding of the interactions between viruses and bacteria in the respiratory tract may enhance insight into the pathogenesis of ARI, and potentially reveal new prevention and treatment strategies. This chapter summarizes the current knowledge on viruses, bacteria and viral-bacterial interactions in childhood ARI and the possible mechanisms by which these interactions may lead to disease.

Acute respiratory infection (ARI), both upper and lower, is a significant cause of childhood morbidity and mortality worldwide. Lower respiratory infection (LRI), including pneumonia, is among the leading causes of childhood mortality worldwide, accounting for close to a million deaths in children under 5 years of age in 2013 [1] . Viruses and bacteria can be detected in most children with ARI. However, both are also frequently detected in asymptomatic children, and hence, the clinical significance of their detection has long been debated. Furthermore, a wide range of prevalence rates are reported for the detection of viruses and bacteria, most likely due to differences in case definitions, diagnostic tools and methodologies used between studies and hence, the epidemiology and etiology of ARI are still not clear. Viruses and bacteria are often co-detected in respiratory samples from children with ARI and their interaction is likely to play a key role in the development of disease. However, the clinical significance of viral-bacterial co-infections and the mechanisms leading to ARI are not well understood. Viruses are identified more frequently than bacteria in children with ARI [2] . Respiratory syncytial virus (RSV) has long been believed to be the most important viral cause of childhood ARI, particularly bronchiolitis, accounting for at least three million hospitalizations and up to 200,000 deaths each year [3] . Human rhinovirus (RV) is the leading cause of upper respiratory infections (URI), but is also an important cause of LRI including pneumonia and bronchiolitis, and accounts for the majority of asthma attacks in children [4, 5] . Advances in molecular methods such as polymerase chain reaction (PCR) in recent years have led to the identification of new viruses and viral species, and the importance of other viruses is now widely acknowledged. Adenovirus, influenza virus, parainfluenza virus, human metapnuemovirus, coronavirus and bocavirus are among other commonly identified respiratory viruses known to contribute to the burden of childhood ARI.

Although less frequently detected than viruses, bacteria were traditionally believed to be the cause of more severe ARI-associated morbidity and mortality, particularly in developing countries [6, 7] . However, traditional diagnostic methods such as blood culture, which are still considered the gold standard in most settings, lack sensitivity and hence, the disease burden attributable to specific bacteria is not well understood. The recent advent of sequencing technologies such as 16S rRNA gene sequencing have led to the discovery of far more diverse respiratory microbial communities than previously recognized [8] . However, advanced diagnostics are largely limited to countries with the most resources and studies detecting a comprehensive range of pathogens remain scarce or non-existent. Streptococcus pneumoniae, Haemophilus influenzae, Moraxella catarrhalis and Staphylococcus aureus are acknowledged as the most common and important bacterial pathogens in childhood ARI, although more recent studies have reported substantial reductions of ARIs due to pneumococcus and Haemophilus influenzae type B (Hib), most likely owing to the widespread introduction of conjugate vaccines [9] .

The respiratory tract is a reservoir of a diverse community of microorganisms, both commensals and potential pathogens [10] . There is growing evidence that synergistic and antagonistic interactions between viruses and bacteria play a key role in the development of ARI. A better understanding of the interactions between viruses and bacteria in the respiratory tract may provide improved insight into the pathogenesis of ARI, and potentially reveal new prevention and treatment strategies. This chapter summarizes the current knowledge on viruses, bacteria and viral-bacterial interactions in childhood ARI and the possible mechanisms by which these interactions may lead to disease.

Respiratory viruses are an important cause of both upper and lower ARI. The burden of viral respiratory infections is greatest in children, with infants and preschool children experiencing an estimated 6-10 viral infections annually [11] . The role of viruses in URI is well established, with 90% of URI caused by viruses compared with only 10% caused by bacteria [12] . Respiratory viruses are also commonly identified in children with LRI, and although their role in LRI is acknowledged, evidence for the clinical significance of respiratory viruses in LRI is still debated. Lung aspiration is considered the gold-standard for sampling since it is obtained directly from the site of infection and would indicate etiological significance for LRI, but is limited due to its invasive nature and rate of complications [11] . Upper respiratory tract samples such as nasopharyngeal aspirates are easily obtained and hence, more frequently used in etiological studies. However, it is argued that viruses identified in the nasopharynx represent only colonization and are not indicative of LRI. Nonetheless, viruses of the upper respiratory tract are more commonly identified in children with LRI than in asymptomatic, healthy children. A 2015 systematic review and meta-analysis of case-control studies of children with and without LRI found evidence for causal attribution of RSV, influenza, parainfluenza virus, human metapnuemovirus and RV in children presenting with LRI compared to asymptomatic or healthy children [13] . However, there was no significant difference in the detection of adenovirus, bocavirus or coronavirus between cases and controls. While RSV and RV are widely acknowledge to be the most common and important viral causes of ARI in children, recent advances in molecular methods have highlighted the significance of other respiratory viruses.

RSV has been considered the most important viral cause of ARI in children, although its contribution is variable [14, 15] . RSV is a seasonal virus that peaks in the cold season in temperate climates and in the rainy season in tropical climates [15] . It affects about 90% of infants and young children by the age of 2 years, with peak rates occurring in infants 6 weeks to 6 months of age [15] . There are two subtypes of RSV, A and B, which circulate concurrently, although some studies have suggested that RSV-A may be more virulent than RSV-B [16] [17] [18] . In 2005, there were an estimated 33.8 million new episodes of RSV-associated ARI worldwide in children less than 5 years of age, with at least 3.4 million RSV-associated ARI hospitalizations, and an estimated 66,000-199,000 deaths, 99% of which occurred in developing countries [3] . Hence, RSV is justifiably seen to be the most important cause of childhood ARI and a major cause of ARI-associated hospital admission [3] . Although RSV has been associated with higher rates of hospitalization than other respiratory viruses in several studies, estimates of RSV-associated ARI incidence and hospitalization are highly variable between and within regions, likely due to methodological differences [15] .

RV is the most commonly identified respiratory virus in both adults and children and was found to be responsible for approximately two-thirds of cases of the common cold [19] . Hence, it is frequently referred to as the "common cold" virus. However, there is now abundant evidence from experimental and observational studies to support the role of RV as a lower respiratory tract pathogen. Early experiment studies with RV suggested that viral replication was optimal at 33 °C (91.4 °F) and was reduced at 37 °C (98.6 °F) and 39 °C (102.2 °F) [20, 21] . However, more recent studies have shown minimal differences in replication capacities at 33 °C (91.4 °F) and 37 °C (98.6 °F) for eight different RV strains, including when viruses were cultured and titrated at the same temperature [22] . Hence, RV is now recognised as an important cause of LRI including pneumonia and bronchiolitis and importantly, accounts for the majority of asthma attacks in children [4, 5] . Several studies worldwide have reported RV as the most common virus identified in children with LRI, with identification rates of up to 63% in some populations [23] .

RV was first isolated and associated with respiratory clinical disease in human in 1956 [24] and by the 1980s, one hundred and one RV-A and RV-B serotypes, known as the reference or prototype had been preserved and distributed by the American Type Culture Collection (ATCC). The retrospective discovery of the third RV species, RV-C, reported in 2006 [25, 26] , led to several new investigations of the prevalence of RV. The majority of these studies in children hospitalized with ARI found that RV-C was the most prevalent RV species and was often associated with more severe illness [5, [25] [26] [27] [28] [29] [30] [31] ].

Adenovirus is another common viral cause of ARI. Adenoviruses are classified into seven species, A to G, and different serotypes have been implicated in different clinical syndromes [32] . While the majority of adenovirus infections present as a mild URI, adenovirus is also known to cause LRI including pneumonia, bronchiolitis and bronchitis [32, 33] . Adenoviruses can also cause gastrointestinal, ophthalmologic, genitourinary and neurological infections. Adenovirus infections are most common during infancy and early childhood and is prevalent in up to 17% of children hospitalized with ARI [32]. Adenovirus is often associated episodes of recurrent wheezing, fever, hypoxia and lengthy hospitalizations [32].

Influenza virus is an important cause of ARI morbidity and mortality, with the highest burden among children less than 5 years of age and adults over 65 years of age [34, 35] . The two main subtypes of influenza virus, A and B, routinely circulate and are responsible for seasonal flu epidemics each year. In the first study to estimate the global incidence of influenza-associated ALRI in children less than 5 years of age, there were an estimated 90 million new cases of influenza episodes, 20 million cases of influenza-associated ARI and one million cases of influenza-associated severe ARI causing 28,000-111,500 deaths worldwide in 2008 [36] . In a more recent systematic analysis of the burden of influenza in paediatric respiratory hospitalizations between 1982 and 2012, influenza was associated with an estimated 870,000 hospitalizations in children less than 5 years of age annually [37].

Human parainfluenza virus (HPIV), first discovered in the late 1950s, is an important cause of ARI in children, accounting for 2-17% of hospitalized cases [38] [39] [40] . There are four types of HPIV, HPIV 1-4. HPIV-1 to HPIV-3 are common causes of ARI in infants and young children, while HPIV-4 is less commonly associated with respiratory illness [41] . Serologic studies have demonstrated that almost all children between 6 and 10 years of age have evidence of past infection, suggesting mild or asymptomatic primary infections [42].

Human metapneumovirus (hMPV) was first isolated from young children with respiratory tract disease the Netherlands in 2001 [43] . Although hMPV was only recently discovered, phylogenetic analysis showed that hMPV has been circulating in humans for at least 50 years [44] . Clinical symptoms of hMPV are similar to those caused by RSV, ranging from URI to severe bronchiolitis and pneumonia [43, 45] . In a study evaluating the burden of hMPV infections in children in the USA, hMPV was detected in 6% of children hospitalized with ARI, 7% of children in outpatient clinics and 7% of children examined in emergency departments, with the greatest burden in children less than 1 year of age [46] . Other studies of hMPV in children hospitalized with ARI have reported prevalences as high as11-25% [47-50].

Human coronavirus is often associated in respiratory illness. Four human coronaviruses (HCoV-229E, HCoV-OC43, HCoV-NL63 and HCoV-HKU1) are endemic in most populations and are associated with mild, self-limiting respiratory illnesses. Another two human coronaviruses, SARS-CoV and MERS-CoV cause severe respiratory syndromes and present a significant threat with their high fatality rates. The four non-severe human coronaviruses are implicated in both URI and LRI, with reported prevalence of approximately 10% in children hospitalized with ARI [51] .

Human bocavirus 1 (hBoV1) was first discovered in 2005 from patients with LRI in Sweden and was the first virus to be discovered by molecular virus screening [52] . Since then, three additional species of HboV, hBoV2, hBoV3 and hBoV4 have been discovered, [53] [54] [55] although these species are found in the gastrointestinal tract and have been associated with gastroenteritis [53] [54] [55] . In contrast, HboV1 is associated with respiratory illness, and more specifically childhood ARI, with reported prevalence ranging from 1.5 to 19% [56] . As with other respiratory viruses, diagnosis of hBoV1 infection is not possible by clinical presentation and for URI common symptoms include common cold-like symptoms and wheezing and for LRI clinical scenarios include pneumonia and bronchiolitis [57] .

The advent of improved molecular methods in recent years has increased the sensitivity in identifying viruses in children with ARI [58] . As a result, identification of multiple viruses in respiratory specimens from children with ARI is frequently reported [59, 60] , with co-infection rates as high as 40-50% [23, 61] . However, the relationship between viral co-infections and severity of ARI in children is not conclusive. Some studies have reported increased risk of ARI hospitalization, increased length of hospital stay and worse clinical outcomes in children with viral coinfections [60, [62] [63] [64] [65] . In contrast, recent systematic reviews and meta-analysis evaluating the relationships between respiratory viral co-infections in children have concluded that viral-coinfections were not associated with ARI severity [66] [67] [68] [69] .

To date, experimental studies of respiratory co-infections are scarce. An in-vitro study examining interactions between RSV and influenza virus demonstrated that growth of RSV was blocked by competitive infection with influenza A virus [70] . In the study by Shinjoh et al., RSV infection produced a higher peak viral load in single infections than in co-infections with influenza virus, if the infections were initiated at the same time [70] . However, if the influenza co-infection was initiated after the RSV infection, influenza growth was suppressed by RSV [70] . The study also demonstrated suppression of the growth of RSV by influenza A infection at the level of viral protein synthesis [70] . Indirect immunofluorescence revealed that a large proportion of infected cells synthesized both RSV and influenza A virus antigens, while scanning electron microscopy demonstrated that influenza A and RSV virions possessing surface antigens specific for each virus were selectively released from dually-infected cells [70] .

In a mathematical model study investigating the dynamics of respiratory viral co-infections, Pinky et al. found that during co-infections, one virus could block another virus by being the first to infect the available host cells and that viral interference through immune response interaction was unlikely [71] . Interestingly, the study found that viral growth rate determines which virus will dominate a simultaneous infection [71] . For example, RV, the fastest-growing virus, reduced replication of the remaining viruses during a co-infection, while parainfluenza virus, the slowest-growing virus is suppressed in the presence of other viruses [71] . The authors of the study suggest that the blocking of one virus infection by the presence of another could be explained through resource competition and this finding has been supported by clinical studies of children with ARI [72, 73] . Canducci et al. found that co-infection of RSV and metapneumovirus in infants with ARI was a protective factor for length of hospital stay and hypoxia, when compared with RSV infection alone [72] . Marguet et al. also found shorter length of hospitalization in infants with RSV and RV co-infection comparing with single RSV infection [73] .

Several other explanations for why identification of multiple viruses do not increase ARI severity have been suggested. One suggestion is that identification of a virus in a respiratory specimen using molecular methods may represent early detection of an infection, asymptomatic carriage, low-virulent infection or prolonged shedding [65] . Therefore, some cases of co-infection may simply be a subject of co-detection. It is also possible that the clinical outcome of viral co-infections is dependent on specific viral combinations, and this may explain the lack of consensus in the literature regarding whether viral co-infections is associated with an increased ARI severity. Many viral co-infection combinations have been reported, as reviewed by Scotta et al., with RSV and RV being the most frequently detected co-infection, although many studies do not even specify the viruses involved in viral co-infections [66] .

Although less frequently detected than viruses, bacteria are also acknowledged to be an important cause of ARI-associated morbidity and mortality. Traditional diagnostic methods such as blood culture, which are still considered the gold standard in most settings, lack sensitivity, and hence the disease burden attributable to specific bacteria is not well understood. Given the absence of a reference standard for the detection of bacterial pathogens, the prevalence of bacteria in ARI is likely to be higher than often reported. S. pneumoniae, H. influenzae, M. catarrhalis and S. aureus are acknowledged as the most common and important bacterial pathogens in childhood ARI, although more recent studies have reported substantial reductions of pneumococcal and Hib disease, most likely owing to the widespread introduction of conjugate vaccines [9] . Culture-independent techniques have also demonstrated that the human microbiome is far greater in extent than previously recognised [74] and that only 1% of all bacteria can be cultured using standard diagnostic methods [8] . However, advanced diagnostics are still rare in most clinical settings and studies detecting a comprehensive range of pathogens remain scarce or non-existent. The lower respiratory tract and lungs were traditionally believed to be sterile and free from bacteria. However, it is now widely accepted that the lungs are constantly exposed to diverse communities of bacteria from the upper respiratory tract and this has been confirmed with the use of culture-independent techniques such as the 16S rRNA gene sequencing. In one of the first studies in the field, Hilty et al. challenged the dogma that the lower respiratory tract is sterile by showing that bronchial tree contains a characteristic microbial flora that differs between health and disease [75] .

The human respiratory tract is home to a diverse community of both commensal and potential pathogenic bacteria that cause respiratory disease. The term "microbiome" was first proposed in 2001 by Joshua Lederberg and is used to describe this "ecological community of commensal, symbiotic and pathogenic microorganisms that literally share our body space." [76] In a balanced state, the respiratory microbiome is believed to play a beneficial role for the human host [77] . However, imbalances in the respiratory microbiome can contribute to the acquisition of new pathogens, carriage of multiple pathogens or interactions among pathogens that lead to respiratory disease. In order to cause respiratory disease, bacterial pathogens must first colonize the nasopharynx. Since the nasopharynx lies between the nose, sinuses, ears, larynx, and the lower respiratory tract, pathogens of the nasopharynx can be the source for both upper and lower respiratory tract infections and hence, plays an important role in both the development of disease and the spread of pathogens [78] . Colonization is believed to be a dynamic and complex microbial process involving acquisition and elimination of species, interactions among microbes and between microbes and the host, and interference by environmental factors [10] . Given these bacteria often co-exist in the same ecological niche; it is likely that highly evolved relationships exist between these bacteria and their interactions with each other play a critical role in the pathogenesis of disease [79] . Furthermore, it is also likely that these species interact with one another even during healthy states [10] .

Streptococcus pneumoniae, or pneumococcus, is a Gram-positive, alpha-hemolytic (under aerobic conditions) or beta-hemolytic (under anaerobic conditions), facultative anaerobic member of the Streptococcaceae family [80] . It is responsible for a range of illnesses including pneumonia, meningitis, bacteremia, otitis media and sinusitis [81] and is the most commonly isolated organism in patients with community-acquired pneumonia [82] . S. pneumoniae colonizes the upper respiratory tract and is part of the normal flora of healthy individuals, particularly children. Although there are over 90 different serotypes, most cases of disease are caused by relatively few serotypes, with the ten most common serotypes accounting for 62% of invasive pneumococcal disease [83] . Since its isolation in 1881, there have been great efforts to treat and prevent S. pneumoniae. Antibiotic treatment for invasive pneumococcal infections typically includes broad-spectrum antibiotics until results of antibiotic sensitivity testing are available. However, emerging antibiotic resistance is a growing concern because of its potential negative impact on the outcome of patients who receive standard antibiotic therapy. Pneumococcal vaccines such as the pneumococcal conjugate vaccine or pneumococcal polysaccharide vaccine are now commonly administered to children globally.

Haemophilus Influenzae is a Gram-negative, coccobacillary, facultative anaerobic pathogenic bacterium belonging to the Pasteurellaceae family [84] . It was first described in 1892 during an influenza pandemic and was mistakenly considered to be the cause of influenza until 1933 when the viral cause of influenza was known [85] . H. influenzae is commonly found in the upper and lower respiratory tract as a commensal but also causes a variety of both invasive infections, such as bacteremia, facial cellulitis, septic arthritis, and meningitis primarily in non-immune children under age 4 years of age, and respiratory tract infections such as pneumonia, acute otitis media, bronchitis, epiglottitis and sinusitis in children and adults [86] . There are six different H. influenzae capsular serotypes, a through f, in addition to nonencapsulated or nontypeable strains. The two most important human pathogens are the capsular serotype b strains and the nontypeable strains which are nonencapsulated. Non-typeable H. influenzae (NTHi) live exclusively in the pharynges of humans and are increasingly recognized as pathogens that cause both localized infections of the respiratory tract (middle ear spaces, sinuses, and bronchi) and systemic infections such as bacteremia and pneumonia [86] .

Moraxella catarrhalis is a Gram-negative, aerobic, oxidase-positive diplococcus belonging to the Moracellaceae family that was first described in 1896 [87] . For most of the past century, M. catarrhalis was regarded as an upper respiratory tract commensal organism. However since the late 1970s, M. catarrhalis has been recognized as an important and common human respiratory tract pathogen [88] . Nasopharyngeal colonization is more prevalent among infants compared with adults, with colonization rates varying between 33 and 100% in infants from different parts of the world [89] [90] [91] . M. catarrhalis is also a common cause of otitis media in infants and children, causing 15-20% of acute otitis media episodes [88] .

Staphylococcus aureus, a Gram-positive coccal bacterium belonging to the Staphylococcaceae family, frequently colonizes the nasopharynx, respiratory tract and skin [92] . It is both a commensal bacterium and a human pathogen, with colonization rates of approximately 30% in the general population, although higher rates are observed in young children and the elderly [92, 93] . S. aureus is also a leading cause of bacteremia and infective endocarditis as well as osteoarticular, skin and soft tissue, pleuropulmonary, and device-related infections [93] .

Several studies have investigated interactions between bacteria, although mostly in experimental and mathematical model studies. In an in-vivo study where H. influenzae was introduced into the nasopharynx of neonatal rats that had or had not been pre-colonized by S. pneumoniae, Margolis et al. reported that H. influenzae density increased when S. pneumoniae was present, suggesting synergism between these bacterial species [94] . However, when these two species were inoculated in the reverse order, inhibition was observed, indicating competition between both species [94] . Another in-vivo study by Lysenko et al. found that both S. pneumoniae and H. influenzae successfully colonized mice when each bacteria species was injected separately [95] . However, when S. pneumoniae was co-colonized with an H. influenzae strain, the density of S. pneumoniae was lower than when inoculated alone, and this was later proved to be fully dependent on complement-and neutrophil-mediated killing of pneumococci [96] . These findings were supported by a large epidemiological study of the 9-valent pneumococcal conjugate vaccine and prevalence of bacterial colonization in HIV-uninfected and HIV-infected children in South Africa that reported inverse associations between S. pneumoniae and S. aureus and between S. aureus and H. influenzae in HIV-uninfected children but not HIV-infected children [97] .

Mathematical models investigating bacterial interactions have produced conflicting results. Using a multivariate random effects model for longitudinal data, Jacoby et al. found a positive association between S. pneumoniae and H. influenzae colonization among Aboriginal and non-Aboriginal children in Australia [98] . In contrast, Pettigrew et al. modeled bacterial colonization in children and reported S. pneumoniae colonization to be negatively associated with colonization by H. influenzae [99] . The study also reported negative associations between S. pneumoniae and S. aureus and between H. influenzae and S. aureus, but interestingly, when H. influenzae was present with M. catarrhalis, the odds of S. pneumoniae colonization increased by more than two-fold [99] . These studies suggest that interactions between bacteria are complex and may shift from negative to positive when additional bacteria species are present.

Epidemiology of Viral-Bacterial Co-infections

Viral-bacterial co-infections are frequently detected in children with respiratory illness, and there is strong evidence for enhanced ARI severity in children during coinfections compared with single infections [100] . While most studies report viral-bacterial co-infection rates ranging from 20 to 50%, rates as high as 66-77% have been observed [70] [71] [72] . The clinical significance of viral-bacterial co-infections and mechanisms that drive these interactions are not well understood. It is difficult to determine the relative importance of individual viruses and bacteria involved in co-infections since both viruses and bacteria can be carried commensally in the respiratory tract and their detection may reflect colonization, rather than infection. Furthermore, it is not possible to distinguish between primary and secondary infections in clinical studies, making it difficult to elucidate the interactions between viruses and bacteria in co-infections. The earliest suggestion that viral infections predispose to secondary bacterial infections has been attributed to French physician R. T. H. Laennec, who observed that the prevalence of pneumonia increased following an influenza epidemic in 1803 [69] . In 1947, British epidemiologist, William Farr, coined the term "excess mortality" to describe the increase in number of deaths during the influenza season that were not caused by influenza itself, and developed the methodology used today to quantitate mortality in influenza epidemics [70] . The association between influenza and secondary bacterial infections came into particular influenza pandemics during the twentieth century, as well as subsequent observational and experimental studies, provided further evidence that viral infections predispose to secondary bacterial infections.

During the 1918 "Spanish flu" pandemic, over 50 million deaths occurred, most of which were not caused directly by influenza alone but rather as a result of secondary bacterial pneumonia [71] . The most frequently identified organisms in sputum, lung and blood samples of infected patients were S. pneumoniae, H. influenzae, Streptococcus pyogenes and S. aureus, and it was believed that the influenza virus acted synergistically with pathogenic bacteria resulting in increased incidence of disease and death [71] . These findings were supported by data from the 1957 "Asian flu" and 1968 "Hong Kong flu" pandemics showing that increased mortality was associated with increased incidence of bacterial pneumonia [72, 73] . The availability of antibiotics effective for secondary bacterial infections was believed to be a key factor for the lower number of deaths during the 1957 and 1968 pandemics compared with 1918. During the 2009 "swine flu" pandemic involving the H1N1 influenza virus, bacterial co-infection was frequently reported in fatal pneumonia cases, with S. pneumoniae being the most frequent bacteria identified [74, 75] .

The best and most studied example of respiratory viral-bacterial co-infections involves influenza virus, and influenza virus-bacterial co-infections has been well described in both adults and children, with clear associations with increased disease severity [101] [102] [103] . However, with the exception of outbreaks, influenza virus is a relatively infrequent viral pathogen compared to other respiratory viruses including RV and RSV. RSV is commonly implicated in viral-bacterial co-infections with reported coinfection rates of up to 17.5-44% in RSV-infected children [104] [105] [106] [107] [108] . In children with severe bronchiolitis studied by Thorburn et al., bacteria were isolated from 42% of lower airway secretions from infants with RSV using culture methods [107] . H. influenzae and S. aureus were the most common bacteria identified and furthermore, the study reported that children with bacterial co-infection were at increased risk for bacterial pneumonia [107] . In serological study of children with communityacquired pneumonia, 39% of children had viral-bacterial co-infections, of which RSV and S. pneumoniae was the most common combination, accounting for 33% of cases [109] . Like influenza, both RSV and S. pneumoniae infection peaks during the winter months [110] and RSV has also been linked to seasonal increases in S. pneumoniae [111] . In a case-control study by Benet et al., co-infection of RSV and S. pneumoniae was more common in cases than in controls but co-infection of RV and S. pneumoniae was not different between cases and controls [112] . RSVbacterial co-infection has also been associated with increased disease severity compared with RSV alone including longer hospital stays and more frequent admission to pediatric intensive care unit [108] and longer ventilator support [105, 107] . However, associations identified in clinical studies are often weak and only occasionally reach statistical significance. Additionally, some studies have reported bacterial co-infection rates below 2% in children with RSV [113] [114] [115] [116] . In a study of infants hospitalized with RSV, bacterial co-infection was found in only 0.6% of children hospitalized for RSV-associated LRI [116] .

RV is also commonly implicated in viral-bacterial co-infections. In a study of children with invasive pneumococcal disease by Techasaensiri et al. 34% of children had a viral co-infection, of which 25% were influenza, and 21% were RV [117] . The study reported that children with viral-coinfections were admitted to the pediatric intensive care unit more frequently and had longer hospital stays than children without viral-coinfections [117] . In a study of children with communityacquired pneumonia by Honkinen et al., viral-bacterial coinfection was identified in 66% of children, of which RV and S. pneumoniae was the most common combination, accounting for approximately 7% of cases [118] . Furthermore, the study reported that all cases of treatment failure had a viral-bacterial co-infection. Lauinger et al. found that among RV-infected children, bacterial co-infections, identified in 8% of children, were associated with increased admission to ICU [119] .

Other respiratory viruses have also been implicated in bacterial co-infections in children, but to a lesser degree. It has been suggested that the pathogenesis of hMPV infection is strongly affected by bacterial co-infection with S. pneumoniae. In a study of children hospitalized with LRI in South Africa, Madhi et al. found that administration of conjugate pneumococcal vaccine reduced the incidence of hMPV infection and the incidence of clinical pneumonia in both HIV positive and negative patients [120] . These findings suggest that a significant proportion of hMPVassociated hospitalizations may be prevented by vaccination with pneumococcal conjugate vaccine. Adenovirus co-infection was identified in 21% of children with invasive pneumococcal disease [117] . Some studies have found associations between overall respiratory virus incidence and bacterial incidence, without distinguishing the specific pathogens involved [110] . While clinical studies confirm that viral-bacterial co-infections are common in children with ALRI, the absence of a control population in most studies makes it difficult to elucidate the clinical significance of co-infections.

The historical context of viral-bacterial co-infections during influenza pandemics have led to a predominantly unidirectional view that primary viral infections increase the development of secondary bacterial infections leading to LRI. Viralbacterial co-infections in children have been described in many clinical studies, with some associations with disease severity. However, in clinical studies, it is difficult to differentiate primary from secondary infections and to elucidate the clinical significance of co-infections. However, several in-vivo and in-vitro experimental model studies have proposed mechanisms to explain interactions between viruses and bacteria. Most mechanisms involve viral facilitation of secondary bacterial infections, for example through disruption of the respiratory epithelium or modulation of innate and adaptive immune responses to decrease bacterial clearance or increase bacterial adherence. However, there is also growing evidence that bacterial infections may promote secondary viral infections, though direct interactions, bacterial interference with antiviral immunity or by synergism or complementation by virulence factors that have similar functions.

The respiratory epithelium is the primary site of host-pathogen encounter in the respiratory tract and the first line of defence against infection [121, 122] . The respiratory epithelium restricts bacterial attachment through mucociliary clearance and maintenance of cell-cell junctions, which restricts access to bacterial receptors [123] . There is evidence that primary viral infections disrupt the respiratory epithelium leading to decreased bacterial clearance. In-vitro studies have shown that cells infected with RV, RSV, adenovirus and influenza led to impairment of mucociliary function and consequent decreased clearance of bacteria including S. pneumoniae and H. influenzae [124] [125] [126] [127] .

Modulation of innate immune cells following a viral-infection is also believed to decrease bacterial clearance in the respiratory tract. Among host innate immune responses, alveolar macrophages are the major cell population in the normal airway and form the first line of defence against respiratory pathogens. A deficiency in alveolar macrophage-mediated phagocytosis following influenza has been reported in several studies. Using a murine-model, Ghoneim et al. showed that influenza infection depleted and induced cell death of alveolar macrophages leading to impaired clearance of S. pneumoniae [128] . Influenza and S. pneumoniae co-infection in mice has been also been shown to result in synergistic stimulation of type 1 interferons (IFNs) leading to impaired recruitment of macrophages and subsequently, increased bacterial colonization [129] . Another murine-model study by Jamieson et al. showed that influenza infection resulted in decreased production of inflammatory cytokines and chemokines through virus-induced glucocorticoid production, reduced recruitment of innate immune cells to the infection site and consequently, a dramatic increase in bacterial burden [130] .

Viral infections of respiratory epithelial cells can also promote bacterial adherence to host cells. In an experimental mouse model study, Hament et al. found that S. pneumoniae adherence to epithelial cells was enhanced by a preceding RSV infection [131] and it was later shown through in-vitro and in-vivo studies that RSV was capable of direct binding to S. pneumoniae [132] . Similarly, Avadhanula et al. showed that respiratory viruses including RSV, HPIV, and influenza virus enhanced adhesion of H. influenzae and S. pneumoniae to primary immortalized cell lines but only RSV and HPIV increased receptor expression for bacteria by primary bronchial epithelial cells and A549 cells [133] . Other studies have shown that RSV virions can bind directly to S. pneumoniae and H. influenzae acting as a direct coupling particle between bacteria and epithelial cells and thereby increasing colonization by, and enhancing, invasiveness of bacteria [124, 134] . There is also evidence that during RSV infection, viral glycoproteins at the host cell surface, act as additional receptions for bacteria adherence [132, 134] . Respiratory viruses can also increase expression of host surface proteins to which bacteria can bind [133, 135, 136] . Host cell receptors for bacterial adherence have also been found to be exposed by viral neuraminidase activities in studies of influenza virus and S. pneumoniae [137, 138] . There is also evidence that viral-mediated epithelia damage can lead to exposure of the basement membrane and additional receptors for bacterial adherence [139, 140] .

Although the majority of evidence for enhanced bacterial adherence during viral respiratory infection comes from studies of influenza virus or RSV, there is growing evidence for the role of RV in bacterial adhesion. In an in-vitro study by Wang et al., nasal epithelial cells were infected with RV, and then S. aureus, S. pneumoniae, or H. influenzae were added to the culture [140] . Compared with RV-uninfected control cells, the adhesion of S. aureus, S. pneumoniae, and H. influenzae increased significantly in RV-infected nasal epithelial cells. In another in-vitro study on the effects of RV infection on the adherence of S. pneumoniae to tracheal epithelial cells, the number of S. pneumoniae adhering to epithelial cells increased after RV infection [141] .

The historical emphasis on influenza-S. pneumoniae co-infections in adults and the unidirectional view that viral infections increase bacterial growth may be less relevant for children. While S. pneumoniae carriage rates are approximately 4% in adults, carriage rates are over 50% in children [142] and up to 80% in children under 5 in developing countries [143] . This supports growing evidence that bacterial infections may promote secondary viral infections, rather than vice versa. In a large, double-blind placebo-controlled trial in infants in South Africa, the 9-valent pneumococcal conjugate vaccine was shown to prevent 31% of pneumonias associated with respiratory viruses in children in hospital, leading to the suggestion that viruses contribute to the pathogenesis of bacterial pneumonia [144] . In another study of healthy children under 2 years of age, Verkaik et al. found that higher seroconversion rates to hMPV were associated with increased nasopharyngeal carriage of S. pneumoniae [145] . Furthermore, well-differentiated normal human bronchial epithelial cells pre-incubated in-vitro with S. pneumoniae resulted in increased susceptibility to infection with HMPV-enhanced green fluorescent protein, suggesting that S. pneumoniae can modulate HMPV infection [145] .

Experimental models often show an increase in influenza virus titres following a bacterial challenge. In one study, influenza viral titres in mice were shown to increase when S. pneumoniae was present [146] . Subsequent mathematical modelling by Smith et al. established that the influenza infection reduced the bacterial clearance ability of alveolar macrophages and that the secondary S. pneumoniae infection enhanced viral release from infected cells [146] . In contrast, in another mouse model study by McCullers et al., influenza infection preceding a pneumococcal challenge primed for pneumonia led to 100% mortality [147] . Interestingly, this effect was specific for viral infection preceding bacterial infection, and reversal of the order of administration led to protection from influenza and improved survival [147] .

Further evidence of bacterial promotion of viral infections has been demonstrated by enhanced RSV and hMPV infection of primary epithelial cells with the addition of bacterial lipopeptides, suggesting that bacteria facilitated viral attachment to host cells [148] . H. influenzae has also been shown to increase airway epithelial cell ICAM-1 and TLR3 expression, leading to enhanced binding of RV and a potentiation of RV-induced chemokine release [149] .

It has also been suggested that viruses might be capable of using their microbial environment to escape immune clearance, highlighting the importance of commensal microbiota in viral infections [150] .

Although viral-bacterial co-infections are common, the clinical significance of coinfections and the mechanisms leading to ARI are yet to be fully clarified. Complex synergistic and antagonistic interactions between viruses and bacteria most likely play a key role in the development of ARI. Viruses and bacteria of the respiratory microbiome may each influence the pathogenicity and consecutive development of infections of the other. Improving knowledge of the interactions between viruses and bacteria may lead to a better understanding of the pathogenesis of ARI and eventually to new prevention and treatment strategies. 

Global, regional, and national causes of child mortality in 2000-13, with projections to inform post-2015 priorities: an updated systematic analysis

The epidemiology of acute respiratory tract infection in young children: comparison of findings from several developing countries. Coordinated Data Group of BOSTID Researchers

Global burden of acute lower respiratory infections due to respiratory syncytial virus in young children: a systematic review and meta-analysis

Rhinovirus and the lower respiratory tract

Association between human rhinovirus C and severity of acute asthma in children

Etiology of acute lower respiratory tract infections in Gambian children: I. Acute lower respiratory tract infections in infants presenting at the hospital

Etiology of acute lower respiratory tract infections in Gambian children: II. Acute lower respiratory tract infection in children ages one to nine years presenting at the hospital

Measurement of in situ activities of nonphotosynthetic microrganisms in aquatic and terrestrial habitats

Communityacquired pneumonia requiring hospitalization among U.S. Children

Viral and bacterial interactions in the upper respiratory tract

Epidemiology of acute lower respiratory disease in children

Systematic review of the treatment of upper respiratory tract infection

Aetiological role of common respiratory viruses in acute lower respiratory infections in children under five years: a systematic review and metaanalysis

Respiratory syncytial virus infection in tropical and developing countries

Respiratory syncytial virus infection

Distribution of respiratory syncytial virus subtypes A and B among infants presenting to the emergency department with lower respiratory tract infection or apnea

Correlation between respiratory syncytial virus genotype and severity of illness

Severity of respiratory syncytial virus infection is related to virus strain

Pathogenesis of rhinovirus common colds in human volunteers

Viruses as precipitants of asthma symptoms II. Physiology and mechanisms

Mechanisms of virus induced exacerbations of asthma

Rhinoviruses replicate effectively at lower airway temperatures

Respiratory viral pathogens associated with lower respiratory tract disease among young children in the highlands of Papua New Guinea

The isolation of a new virus associated with respiratory clinical disease in humans

MassTag polymerase-chainreaction detection of respiratory pathogens, including a new rhinovirus genotype, that caused influenza-like illness in New York State during

Frequent detection of human rhinoviruses, paramyxoviruses, coronaviruses, and bocavirus during acute respiratory tract infections

Cinical and molecular epidemiology of human rhinovirus C in children and adults in Hong Kong reveals a possible distinct human rhinovirus C subgroup

A recently identified rhinovirus genotype is associated with severe respiratory-tract infection in children in Germany

A novel group of rhinoviruses is associated with asthma hospitalizations

A 1-year experience with human metapneumovirus in children aged <5 years

Update on human rhinovirus and coronavirus infections

Cloning of a human parvovirus by molecular screening of respiratory tract samples

A novel bocavirus associated with acute gastroenteritis in Australian children

A new bocavirus species in human stool

Human bocaviruses are highly diverse, dispersed, recombination prone, and prevalent in enteric infections

Human bocavirus

Identification of new respiratory viruses in the new millennium

Evaluation of multiple commercial molecular and conventional diagnostic assays for the detection of respiratory viruses in children

Multiple simultaneous viral infections in infants with acute respiratory tract infections in Spain

Correlation of viral load of respiratory pathogens and co-infections with disease severity in children hospitalized for lower respiratory tract infection

High rate of viral identification and coinfections in infants with acute bronchiolitis

The impact of dual viral infection in infants admitted to pediatric intensive care unit associated with severe bronchiolitis

Viruses in community-acquired pneumonia in children aged less than 3 years old: high rate of viral coinfection

The burden of single virus and viral coinfections on severe lower respiratory tract infections among preterm infants: a prospective birth cohort study in Brazil

Clinical utility of PCR for common viruses in acute respiratory illness

Respiratory viral coinfection and disease severity in children: A systematic review and metaanalysis

Systematic review and meta-analysis of respiratory viral coinfections in children

Single and multiple respiratory virus infections and severity of respiratory disease: a systematic review

Clinical disease severity of respiratory viral co-infection versus single viral infection: a systematic review and metaanalysis

In vitro growth profiles of respiratory syncytial virus in the presence of influenza virus

Coinfections of the respiratory tract: viral competition for resources

Two-year prospective study of single infections and co-infections by respiratory syncytial virus and viruses identified recently in infants with acute respiratory disease

In very young infants severity of acute bronchiolitis depends on carried viruses

The human microbiome project

Disordered microbial communities in asthmatic airways

Ome sweet' omics-a genealogical treasury of words

What are the consequences of the disappearing human microbiota?

Dynamics of nasopharyngeal colonization by potential respiratory pathogens

Microbial interactions in the respiratory tract

Sherris medical microbiology

Streptococcus pneumoniae: description of the pathogen, disease epidemiology, treatment, and prevention

The aetiology and antibiotic management of community-acquired pneumonia in adults in Europe: a literature review

Pneumococcal serotypes and their clinical relevance

Pasteurellaceae: biology, genomics and molecular aspects

Discovery and characterization of the 1918 pandemic influenza virus in historical context

What the pediatrician should know about non-typeable Haemophilus influenzae

Branhamella catarrhalis: epidemiology, surface antigenic structure, and immune response

Moraxella catarrhalis, a human respiratory tract pathogen

Nasopharyngeal colonization during the first year of life

Epidemiology of Moraxella catarrhalis in children during the first 2 years of life: relationship to otitis media

Bacterial colonization of the nasopharynx predicts very early onset and persistence of otitis media in Australian aboriginal infants

The role of nasal carriage in Staphylococcus aureus infections

Staphylococcus aureus infections: epidemiology, pathophysiology, clinical manifestations, and management

The ecology of nasal colonization of Streptococcus pneumoniae, Haemophilus influenzae and Staphylococcus aureus: the role of competition and interactions with host's immune response

The role of innate immune responses in the outcome of interspecies competition for colonization of mucosal surfaces

Within-host competition drives selection for the capsule virulence determinant of Streptococcus pneumoniae

Long-term effect of pneumococcal conjugate vaccine on nasopharyngeal colonization by Streptococcus pneumoniae-and associated interactions with Staphylococcus aureus and Haemophilus influenzae colonization-in HIV-Infected and HIV-uninfected children

Modelling the cooccurrence of Streptococcus pneumoniae with other bacterial and viral pathogens in the upper respiratory tract

Microbial interactions during upper respiratory tract infections

Viral bacterial co-infection of the respiratory tract during early childhood

Severe Pneumococcal pneumonia in previously healthy children: the role of preceding influenza infection

Streptococcus pneumoniae coinfection is correlated with the severity of H1N1 pandemic influenza

Community-acquired respiratory coinfection in critically ill patients with pandemic 2009 influenza A(H1N1) virus

Pulmonary and systemic bacterial co-infections in severe RSV bronchiolitis

Concurrent bacterial infection and prolonged mechanical ventilation in infants with respiratory syncytial virus lower respiratory tract disease

Risk of bacterial infection in previously healthy respiratory syncytial virus-infected young children admitted to the intensive care unit

High incidence of pulmonary bacterial co-infection in children with severe respiratory syncytial virus (RSV) bronchiolitis

Risk of concurrent bacterial infection in preterm infants hospitalized due to respiratory syncytial virus infection

Etiology of childhood pneumonia: serologic results of a prospective, population-based study

Association of invasive pneumococcal disease with season, atmospheric conditions, air pollution, and the isolation of respiratory viruses

Influence of pneumococcal vaccines and respiratory syncytial virus on alveolar pneumonia

Etiology and factors associated with pneumonia in children under 5 years of age in Mali: a prospective case-control study

Concurrent serious bacterial infections in 2396 infants and children hospitalized with respiratory syncytial virus lower respiratory tract infections

Prevalence of serious bacterial infections in febrile infants with respiratory syncytial virus infection

Bacteraemia and antibiotic use in respiratory syncytial virus infections

Risk of secondary bacterial infection in infants hospitalized with respiratory syncytial viral infection

Viral coinfections in children with invasive pneumococcal disease

Viruses and bacteria in sputum samples of children with community-acquired pneumonia

Patient characteristics and severity of human rhinovirus infections in children

Pneumococcal coinfection with human metapneumovirus

Epithelial cells and airway diseases

The airway epithelium: soldier in the fight against respiratory viruses

Compromised defenses: exploitation of epithelial responses during viral-bacterial co-infection of the respiratory Tract

Respiratory syncytial virus increases the virulence of Streptococcus pneumoniae by binding to penicillin binding protein 1a. A new paradigm in respiratory infection

Influenza virus infection decreases tracheal mucociliary velocity and clearance of Streptococcus pneumoniae

Synergistic effect of adenovirus type 1 and nontypeable Haemophilus influenzae in a chinchilla model of experimental otitis media

Rhinovirus disrupts the barrier function of polarized airway epithelial cells

Depletion of alveolar macrophages during influenza infection facilitates bacterial superinfections

Synergistic stimulation of type I interferons during influenza virus coinfection promotes Streptococcus pneumoniae colonization in mice

Influenza virus-induced glucocorticoids compromise innate host defense against a secondary bacterial infection

Enhanced adherence of Streptococcus pneumoniae to human epithelial cells infected with respiratory syncytial virus

Direct binding of respiratory syncytial virus to pneumococci: a phenomenon that enhances both pneumococcal adherence to human epithelial cells and pneumococcal invasiveness in a murine model

Respiratory viruses augment the adhesion of bacterial pathogens to respiratory epithelium in a viral species-and cell type-dependent manner

Nontypeable Haemophilus influenzae and Streptococcus pneumoniae bind respiratory syncytial virus glycoprotein

Streptococcus pneumoniae anchor to activated human cells by the receptor for platelet-activating factor

Transcriptional adaptation of pneumococci and human pharyngeal cells in the presence of a virus infection

Role of neuraminidase in lethal synergism between influenza virus and Streptococcus pneumoniae

Contribution of vaccine-induced immunity toward either the HA or the NA component of influenza viruses limits secondary bacterial complications

Adherence of type I Streptococcus pneumoniae to tracheal epithelium of mice infected with influenza A/PR8 virus

Rhinovirus enhances various bacterial adhesions to nasal epithelial cells simultaneously

Effects of rhinovirus infection on the adherence of Streptococcus pneumoniae to cultured human airway epithelial cells

Nasopharyngeal carriage of Streptococcus pneumoniae by adults and children in community and family settings

Carriage of Streptococcus pneumoniae and other respiratory bacterial pathogens in low and lower-middle income countries: a systematic review and meta-analysis

A role for Streptococcus pneumoniae in virus-associated pneumonia

Streptococcus pneumoniae exposure is associated with human metapneumovirus seroconversion and increased susceptibility to in vitro HMPV infection

Kinetics of coinfection with influenza A virus and Streptococcus pneumoniae

Lethal synergism between influenza virus and Streptococcus pneumoniae: characterization of a mouse model and the role of platelet-activating factor receptor

The synthetic bacterial lipopeptide Pam3CSK4 modulates respiratory syncytial virus infection independent of TLR activation

influenzae potentiates airway epithelial cell responses to rhinovirus by increasing ICAM-1 and TLR3 expression

Successful transmission of a retrovirus depends on the commensal microbiota