RESPIRATORY VIRUS ALTERS ASYMPTOMATIC CARRIAGE OF KNOWN BACTERIAL PATHOGENS
Consistent with our first hypothesis, viral infection frequently has been associated with carriage of common pneumonia pathogens. In a cross-sectional analysis of aboriginal and non-aboriginal children in Western Australia, Jacoby et al. [18] observed positive associations between rhinovirus and S. pneumoniae, H. influenzae and M. catarrhalis and a positive association between adenovirus and M. catarrhalis in the nasopharynx. In a US study, children with a viral URT co-infection not associated with otitis media had a higher prevalence of nontypeable H. influenzae and M. catarrhalis relative to healthy children. Furthermore, children with viral co-infection associated with acute otitis media had an increased prevalence of S. pneumoniae, nontypeable H. influenzae and M. catarrhalis but a decreased prevalence of α-hemolytic Streptococci [77]. van den Bergh et al. [26] assessed the prevalence of 20 respiratory viruses and the main commensal bacteria in the nasopharynx of 433 healthy Dutch children aged 6–24 months. In their study, rhinovirus was positively associated with S. pneumoniae and H. influenzae, RSV was positively associated with H. influenzae, coronaviruses and adenovirus were positively associated with M. catarrhalis, and influenza virus was positively associated S. aureus (Fig. 2). However, as the associations found in the above-mentioned studies are based on cross-sectional analyses, we cannot determine whether viruses influenced carriage structure, bacterial carriage influenced host susceptibility to viruses or if bidirectional interactions occurred. Prospective studies are required to resolve this temporal ambiguity.
Figure 2.  Network of interactions between virus and bacteria in the upper respiratory tract. Figure 1A in van den Bergh et al. [26] used under the Creative Commons Attribution License. Green lines indicate synergistic associations and red lines indicate antagonistic associations. Solid lines indicate associations with P < 0.01 and dashed lines indicate associations with P between 0.01 and 0.05 for associations between species. enterovirus (EV), H. influenzae (HI), human adenovirus human (HAdV), bocavirus (HBoV), human coronavirus (HCov), human parechovirus (HPeV), human rhinovirus (HRV), influenza virus (IV), M. catarrhalis (MC), S. aureus (SA), S. pneumoniae (SP) and WU polyomavirus (WUPyV)   Although the impact of the host microbiota on viral infections is an important consideration (reviewed by Wilks et al. [78]), the majority of in vivo experiments pertaining to virus–bacterial interactions in the URT focus on the role of viruses on the host microbiota. The results of these studies suggest that viruses can alter carriage structure by promoting the colonization of certain commensals. In both animal models and human adults, infection with influenza A virus showed increased colonization by S. pneumoniae and H. influenzae in the URT [79–83]. Similarly, infecting rats and chinchillas with RSV led to increased colonization by nontypeable H. influenzae [84, 85]. Collectively, epidemiologic studies and laboratory experiments suggest that the introduction of a virus to the URT niche can substantially alter the bacterial community present [26].