In contrast to viral infections, which are known to be associated with wheezing in young children, bacterial infections have long been considered of no importance in asthma and wheezing, and antibiotic therapy is not recommended in current asthma treatment guidelines [17–20]. However, several studies suggest that bacteria might play a role in wheezing in young children [9, 10, 17, 19, 20]. Moreover, it is increasingly appreciated that the human body is home to an extended microbiome, with asthmatic children harbouring different bacteria and a less diverse mix compared to healthy children . Characterisation of the predominant bacteria in wheezing children is essential to understand the pathophysiology of persistent wheezing.
Recently, tracheobronchomalacia has been shown to be highly associated with bronchial inflammation and bacterial infection in children with persistent respiratory symptoms . Nevertheless, it remains unclear whether bacterial infection is simply a consequence of impaired mucociliary clearance or whether it is an independent inflammatory stimulus in persistent wheezers.
We hypothesized that bacterial infection might indeed be an independent inflammatory stimulus in persistent preschool wheezers, and chose to study a selected group of patients without structural anomalies of the conductive airways.
In our study, we found that bronchial bacterial infection is common in persistent preschool wheezers without apparent reasons for impaired mucociliary clearance, and that H .influenzae was the predominant aerobic pathogen isolated, followed by S. pneumoniae and M. catarrhalis. We also found frequent bacterial co-infection involving these 3 pathogens. Our findings corroborate observations of other authors, who also described the predominance of this bacterial triad in children with persistent or recurrent wheezing [9, 10, 19, 20, 23, 24].
The frequent upper respiratory tract colonisation with these aerobic bacteria in children, may provide difficulties to discriminate between contamination and true bronchial infection . However, we consider our results to reflect true bronchial infection, as we applied procedural methods, which were shown to limit strongly the risk for contamination [13, 20, 26, 27]. In particular, we chose to use a cut-off of ≥104 cfu/ml for bacterial BAL-culture, although in adults a cut-off of ≥103 cfu/ml has demonstrated to be discriminative [28, 29]. The applied cut-off was also used previously by others, studying children with chronic bronchitis . Moreover, although FOB + BAL were only performed once in the majority of patients, our results probably reflect chronic bronchial infection as all our samples were collected during stable clinical state.
The presence of a significant increased neutrophilic inflammation in BAL-fluid of those patients in which NTHi was cultured, as compared to patients with a negative BAL-culture, supports the hypothesis that bacterial infection is indeed an inflammatory stimulus in children with persistent wheezing.
H. influenzae was the predominant aerobic bacterial pathogen, isolated in 30.3 % of our patients. Although previously other authors have had similar results, isolated strains of H. influenzae were never further characterised [9, 10, 20, 23]. To our knowledge this is the first report that associates bronchial infection with non-typeable H. influenzae (NTHi) in young persistant wheezers. Identification of NTHi as the predominant pathogen further supports our hypothesis that persistent wheezers present an underlying inflammation induced by chronic bacterial infection. Chronic NTHI infection, resulting from the ability of NTHI to persist in the lower airways through several escape mechanisms from the host immune responses, has previously been reported by others [31–33]. Moreover, in the lower airways of patients with chronic obstructive lung disease, NTHI has been demonstrated to be a major inflammatory stimulus [31, 34].
The poor response of symptoms to an in-vitro susceptibility guided, adequately dosed antibiotic treatment that we found, is similar to poor response findings in 2 other disease entities involving H. influenzae as a predominant pathogen in children; recurrent and unresponsive otitis media and protracted bacterial bronchitis [35–38]. In the middle ear, biofilm formation has been demonstrated to be one of the major mechanisms for persistence and provides an explanation for unresponsiveness to antibiotic therapy [39, 40]. Moreover, biofilm formation and intracellular infection are well established mechanisms of NTHi for persistence in the lower airways [31–33]. Although our study does not provide evidence of biofilm formation in the lower airways of young unresponsive wheezers, our findings of poor response to and frequent relapse after a short course of antibiotic treatment are suggestive for involvement of biofilm. Biofilm formation should be addressed in future studies, because its presence may have consequences for treatment.
The strength of our study is the exhaustive microbiological evaluation with use of semi-quantitative BAL-cultures, multiplex-PCR (mPCR) for viral and atypical agents, and thorough characterisation of the isolated H. influenzae strains. However, we may have missed some cases of viral infection because human rhinovirus was not included in the applied PCR-protocol.
The major weakness of our study is its retrospective character and the absence of an age-matched control group. However, reports, from prospective conducted studies without profound characterisation of the isolates, describing association between wheezing and bacterial infection support our findings [10, 19, 22]. The absence of a control group is considered a minor weakness, as previous publications have demonstrated the limited risk for contamination of BAL when technical precautions are taken [13, 20, 26, 27]. Moreover, performance of FOB + BAL is considered unethical in healthy children [10, 12].
Another limitation is the small patient cohort. Due to the multiple exclusion criteria applied, the patients included in the analysis represented only 1,2 % of children seen in our clinic because of persistent respiratory symptoms. Nevertheless, analysis of this highly selected population without structural causes of impaired mucus clearance, enabled us to evaluate the association between bacterial bronchial infection, neutrophilic inflammation and symptoms resistant to ICS therapy.