Article Text
Abstract
Background Respiratory viral infections precipitate exacerbations of chronic respiratory diseases such as asthma and chronic obstructive pulmonary disease though similar data in non-cystic fibrosis (CF) bronchiectasis are missing. Our study aimed to determine the point prevalence of viruses associated with exacerbations and evaluate clinical and investigational differences between virus-positive and -negative exacerbations in children with bronchiectasis.
Methods A cohort of 69 children (median age 7 years) with non-CF bronchiectasis was prospectively followed for 900 child-months. PCR for 16 respiratory viruses was performed on nasopharyngeal aspirates collected during 77 paediatric pulmonologist-defined exacerbations. Clinical data, systemic (C reactive protein (CRP), IL-6, procalcitonin, amyloid-A, fibrinogen) and lung function parameters were also collected.
Findings Respiratory viruses were detected during 37 (48%) exacerbations: human rhinovirus (HRV) in 20; an enterovirus or bocavirus in four each; adenoviruses, metapneumovirus, influenza A virus, respiratory syncytial virus, parainfluenza virus 3 or 4 in two each; coronavirus or parainfluenza virus 1 and 2 in one each. Viral codetections occurred in 6 (8%) exacerbations. HRV-As (n=9) were more likely to be present than HRV-Cs (n=2). Children with virus-positive exacerbations were more likely to require hospitalisation (59% vs 32.5% (p=0.02)) and have fever (OR 3.1, 95% CI 1.2 to 11.1), hypoxia (OR 25.5, 95% CI 2.0 to 322.6), chest signs (OR 3.3, 95% CI 1.1 to 10.2) and raised CRP (OR 4.7, 95% CI 1.7 to 13.1) when compared with virus-negative exacerbations.
Interpretation Respiratory viruses are commonly detected during pulmonary exacerbations of children with bronchiectasis. HRV-As were the most frequently detected viruses with viral codetection being rare. Time-sequenced cohort studies are needed to determine the role of viral–bacterial interactions in exacerbations of bronchiectasis.
- Bronchiectasis
- Exacerbation
- Virus
- Biomarkers
- Children
Statistics from Altmetric.com
What is already known on this topic?
-
Respiratory viral infections precipitate exacerbations of chronic respiratory diseases such as asthma and chronic obstructive pulmonary disease.
-
Similar data in non-cystic fibrosis (CF) bronchiectasis are missing.
-
Pulmonary exacerbations increase morbidity and are associated with progressive deterioration in lung function and poor quality of life.
What this study adds?
-
This prospective cohort study reports the prevalence and outcomes of respiratory virus related exacerbations in bronchiectasis.
-
Among children with non-CF bronchiectasis, respiratory viruses cause pulmonary exacerbations and are associated with worse clinical outcomes.
-
Human rhinoviruses-As were the most frequently detected viruses with viral codetection being rare.
Introduction
Non-cystic fibrosis (non-CF) bronchiectasis is being increasingly diagnosed and recognised as an important contributor to chronic lung disease in both adults and children in developing1 and developed countries.2 Pulmonary exacerbations increase morbidity,3 and are associated with progressive deterioration in lung function4 and poor quality of life,5 but little research data are available. Indeed, though clinically suspected, there are no published prospective data on viral triggers of pulmonary exacerbations of non-CF bronchiectasis. In other chronic respiratory diseases (such as asthma,6 ,7 CF8 ,9 and chronic obstructive pulmonary disease (COPD)),10–12 respiratory viruses are known to be common triggers of exacerbations. In two separate retrospective studies on children with non-CF bronchiectasis, we reported that 21% of upper airway samples during exacerbation and 12% of lower airway samples during stable state were positive for respiratory viruses.3 ,13
The importance of understanding the role of viruses in bronchiectasis exacerbations is manifold. First, the effect of viral infection could range from trivial changes in clinical status to precipitating inflammatory events that lead to sustained and severe deterioration as seen in COPD.10 ,12 Published studies suggest that nearly half of the exacerbations in COPD11 and CF8 are directly or indirectly related to viral agents with human rhinoviruses (HRVs) being found most frequently. Lack of data in children or adults with bronchiectasis limits comparison with these studies. Second, antibiotic use during an exacerbation of bronchiectasis is usually based on clinical criteria and the lack of viral data potentially leads to overuse or inappropriately timed use of antibiotics. Third, although influenza vaccination is routinely recommended in bronchiectasis, the absence of direct data is a potential obstacle promoting its use. Fourth, identifying the contributions of various respiratory virus types (including newly-identified respiratory viruses) in bronchiectasis exacerbations is potentially important as antiviral agents (such as neuraminidase inhibitors and monoclonal antibodies) are becoming increasingly available.14 Last, knowledge gained in this area will potentially lead to understanding new patho-physiological mechanisms relevant for bronchiectasis, as has been shown for asthma.7
In addition to the knowledge gaps identified above, there are limited published data on viral-associated exacerbations in bronchiectasis and markers of systemic inflammation. These potential systemic markers include procalcitonin (PCT), interleukin 6 (IL-6), serum amyloid A (SAA), fibrinogen and C reactive protein (CRP) though only PCT has been reported to predict response to antibiotic therapy in respiratory tract infections.15
To address these gaps, we conducted a prospective cohort study of children with non-CF bronchiectasis. We aimed to: (a) determine the point prevalence of respiratory viruses (including newly-identified viruses) associated with pulmonary exacerbations and (b) evaluate clinical and investigational differences between virus-positive and -negative exacerbations in children with bronchiectasis.
Methodology
Subjects
Children (aged <18 years) with radiology-based diagnosis of bronchiectasis without CF presenting at the Royal Children's Hospital were prospectively enrolled from 1 February 2008 to 31 January 2010 and followed until 31 July 2010. This was part of a larger prospective cohort study to delineate a validated definition of pulmonary exacerbation in children with non-CF bronchiectasis.16 All children had bronchiectasis diagnosed by high resolution CT scan of the chest and had a sweat chloride less than 35 meq/L. Parents were asked to contact the primary investigator (NK) when their child was suspected of experiencing an exacerbation. Children were reviewed at the hospital clinic every 3 months and when suspected of experiencing an exacerbation. Informed parental consent was obtained at the time of enrolment. The Queensland Health Children's Health Services Ethics Committee approved the study (Approval ID-2008/038).
At the clinic visits, children undertook a standardised assessment.16 Nasopharyngeal aspirate (NPA) was collected for viral studies and blood was tested for potential markers of exacerbations (SAA, IL-6 and PCT), full blood count with differential white cell counts and acute phase reactants (CRP and fibrinogen). The decision to hospitalise the child was made by the treating pulmonologist.
Virus identification
The NPA was assessed using an extended respiratory viruses PCR screen that detected influenza viruses A and B; human parainfluenza virus (HPIV) 1–4; adenoviruses; human respiratory syncytial virus (RSV); human metapneumovirus (hMPV); HRV A, B and C; human coronavirus (HCoV); human bocavirus (HBoV); and human enterovirus. NPA was assessed for only the routine eight respiratory viruses (influenza viruses A and B, HPIV 1–3, adenoviruses, RSV and hMPV) during the stable state, as described previously17 (please refer to online supplementary file for details on methodology of viral assays).
Statistical analysis
Descriptive statistics for continuous data were expressed as median (IQR). Mann–Whitney U test was used to compare continuous variables and χ2 or Fisher exact test for categorical variables. Logistic regression was undertaken to derive a model of factors associated with virus-positive exacerbation state. The dependent variable was NPA viral PCR-positive exacerbation state versus virus-negative exacerbation state. All independent variables were dichotomised.16 Factors found significant at p<0.2 in univariate analysis were included for multiple logistic regression. Multiple regression was done separately for clinical and investigational models with age included in the model as a clinically significant variable. A p value<0.05 was considered significant in the multiple regression model. All analysis was done using SPSS 13.0.
Results
A total of 69 children (33 boys) were prospectively enrolled and followed for 900 child-months. The median (IQR) age at enrolment was 7 (3.8, 10.9) years and that at diagnosis was 5.7 (3.1, 8.1) years. In all, 59 (85%) children were Caucasian, four were Indigenous, four Asian and two Maori. Of these, 18 (27%) children had unilobar bronchiectasis; 30 (43.5%) of the children had ‘idiopathic bronchiectasis’ and 12 had primary immunodeficiency.16 In the 69 children followed up for a median (IQR) duration of 13 (8–19) months, a total of 81 pulmonary exacerbations were included from 65 children; four children did not have any pulmonary exacerbation during the follow-up. Extended viral assay was conducted during 77 of the 81 exacerbations. In the remaining four children, NPAs were only tested for classical respiratory viruses. NPAs were tested during the stable state for the routine (eight viruses) screen in 62 of the 69 children.
From the 77 exacerbations with complete data, at least one respiratory virus was detected in 37 (48%) and viral codetections occurred in 6 (8%) exacerbations. Fever was reported during 15 (19%) exacerbations and coryzal symptoms (running nose and/or sore throat) from 38 (49%). Only three exacerbations had fever without coryzal symptoms, one of which was positive for non-typable HRV and the other two were virus-negative. Of the 41 exacerbations with fever and/or coryzal symptoms, 34 (83%) were virus-positive. Only 3 (8%) of the 37 virus-positive exacerbations did not have clinical symptoms of a viral infection. Further, 22 (59%) of the virus-positive exacerbations required hospitalisation compared with 32.5% of virus-negative exacerbations (p=0.02).
The point prevalence of the viruses detected is described in table 1. HRVs were the most frequently detected viruses with HRV-A (nine of 20) significantly more common than HRV-C (two of 20) viruses; difference between proportions of 0.35 (95% CI 0.1 to 0.61). The six exacerbations with viral codetection were: HRV-B with HBoV (n=2); RSV with HMPV (n=2); and one each of HRV with HCoV and HBoV with adenovirus. None of the children with more than one exacerbation included for analysis had the same virus detected during the subsequent exacerbation.
Four children were positive for at least one respiratory virus during the stable state. Of these, two were adenovirus-positive and one each was positive for HPIV 3 virus and RSV. The child who had adenovirus detected during the stable state had an adenoviral-positive exacerbation approximately 8 weeks prior to the stable state assessment.
On univariate logistic regression analysis for factors associated with virus-positive exacerbation (table 2), significant variables (at p<0.2) were cough severity, sputum production, chest pain, fever, chest signs, hypoxia, raised CRP and IL-6. However, in a multiple regression model after adjusting for age and other factors found significant on univariate analysis, the clinical factors that remained significant were presence of fever, chest signs and hypoxia. The only investigational factor that remained significant on multiple regression was raised CRP (table 3).
Of the 77 exacerbations, 16 occurred in the 11 children with primary immunodeficiency. Of these 16 exacerbations, seven were virus-positive (44%) compared with 30 (49%) in the children without primary immunodeficiency (p=0.72). Similarly, the seven virus-positive exacerbations in this group did not differ in severity (57% needed hospital admission in the immunodeficiency group with virus-positive exacerbation compared with 60% in the non-immunodeficiency group; p=0.88) or inflammatory markers compared with the other subgroups (p range 0.64–0.9).
Discussion
The current lack of data on respiratory viruses associated with pulmonary exacerbation is a clinical and research gap in children with non-CF bronchiectasis. Using PCR for a large panel of classical and newly-identified respiratory viruses, we identified respiratory viruses in the NPAs at the onset of 48% of the 77 paediatric pulmonologist-defined exacerbations in a prospective cohort of 69 children with non-CF bronchiectasis. HRV and HPIV were the most commonly detected viruses with viral codetections being uncommon. Virus-positive exacerbations were more likely to be associated with fever, hypoxia, chest signs and hospitalisation. CRP was significantly higher during virus-positive exacerbations but no other systemic marker differentiated viral from non-viral exacerbations.
This is the first study that has prospectively examined the prevalence of viruses in patients (children or adults) with bronchiectasis. Even with the use of a large panel of classical and newly-identified respiratory viruses, only 48% of exacerbations were associated with a virus-positive state. While this is comparable with CF8 (46%) and COPD11 (51%), it is in contrast to paediatric cohorts of asthma (60%–80%).7 ,18
The mechanisms by which viruses cause exacerbations of a chronic pulmonary disease range from direct infection of the lower respiratory tract19 to virus induced suppression of antimicrobial peptides like elafin resulting in secondary bacterial infections.20 Other potential mechanisms include virus induced immune modulation,21 neural reflex responses and lower airway inflammation.12 Experimental studies in COPD have shown that HRV infection may impair interferon production and induce a neutrophilic inflammation causing pulmonary exacerbations22 though a similar effect on interferon has not been consistently reported.23 HRV has also been reported to increase 5-lipoxygenase and cyclooxygenase-2 in the airways of non-atopic subjects.24 These inflammatory mediators may result in an increase in bronchial wall thickness and luminal exudates and these may be responsible for the clinical features of exacerbations.
In our study, HRV was the most frequently detected virus, consistent with exacerbation data in CF,9 asthma6 ,18 and COPD.11 Our study is the first in bronchiectasis to prospectively analyse the distribution of HRV species. In contrast to other studies with available HRV typing (eg, paediatric cohorts in asthma6), we found that HRV-A viruses were significantly more prevalent in bronchiectasis than HRV-C. The role of HRV-C in exacerbation of chronic respiratory diseases has only recently been described6 ,25 and this species is increasingly found to be the most prevalent subtype. Whether this difference in distribution of HRV in our cohort is disease specific or just reflects seasonal or year to year variation remains unknown. Further, in contrast to paediatric asthma cohorts, codetection of viruses was rare (42%6 vs 8%). Additionally, we have also described several rarely sought or newly-identified viruses such as HPIV 4, HBoV and HCoVs not previously reported in bronchiectasis. HPIV 4 has been reported as a cause of severe respiratory infections in previously healthy children26 and has been reported in CF exacerbations.8 Our study likewise found HPIV 4 in bronchiectasis exacerbations but further studies are required to define its role in disease severity and progression. It is also unclear if these viruses are pathogens or passengers as a recent study in asymptomatic children has described presence of viruses, detected by molecular techniques, in more than half of the studied children.27
Children with pulmonary exacerbations associated with a virus-positive state were clinically more unwell (with hypoxia and fever) and more likely to require hospitalisation. Further, a higher proportion of virus-positive exacerbations had coryzal symptoms. These findings are similar to CF8 and COPD studies10 ,11 and indirectly implicate the viruses in pathogenesis of the exacerbation state. Although Wat et al8 report no difference in the peak expiratory flow between virus-positive and -negative exacerbation in 71 children with CF, higher lung function decline has been reported in viral exacerbations of COPD.12 This suggests that respiratory viruses might be important in causing and determining the severity of exacerbations.
The finding that systemic features (such as fever and chest signs) were more common in the virus-positive group indicates that in the absence of a viral illness, these are uncommon features of an exacerbation in bronchiectasis (only 10% children with virus-negative exacerbation had fever). This is highly relevant as, in countries like Australia, some general practitioners and paediatricians are reluctant to start antibiotics in the absence of systemic features such as chest signs and fever. Such an approach potentially undertreats these exacerbations because in our experience,16 bronchiectasis exacerbations usually do not present with fever, with the most important symptoms being an increased frequency of cough and/or change in character of cough. Further, the role of antibiotics in virus-positive exacerbation is unknown, especially in the absence of concurrent bacterial data. Since concomitant lower airway bacterial data are usually unobtainable during exacerbations in young non-expectorating children with bronchiectasis, a pragmatic approach of treating these exacerbations with antibiotics covering the usual organisms such as Streptococcus pneumonia, Haemophilus influenzae and Moraxella catarrhalis may be justified.13 Future studies on viral–bacterial interactions are needed to answer this question.
Limitations of our study include absence of extended-panel viral data during the stable state. In the absence of extended-panel stable state viral data, causality of viral agents, especially HRV, cannot be concluded in our cohort. Also, we did not have bacterial data to determine the contribution of bacterial co-infection. In adults with COPD, it has been shown that viral infection with HRV precipitates secondary bacteria infection.20 ,22 In our cohort, the children were young and could not expectorate; obtaining lower airways specimens during each exacerbation was not feasible. Moreover, the analysis of potential biomarkers was opportunistic. These did not take into account bacteria data and are also likely not powered to detect a significant difference between these virus-positive and -negative groups.
Despite the limitations above, the novelty of our study is manifold. First, ours is the first prospective study that has sought viruses in any non-CF bronchiectasis cohort. Second, the panel of viruses we examined using state of the art molecular techniques included classical and newly-identified respiratory viruses. Using this panel, we found previously undescribed viruses associated with exacerbations. Third, we found that HPIVs and HRV-As may be particularly important in bronchiectasis exacerbations although their exact significance needs to be further defined in a larger multicentre multiyear cohort. Notably, influenza A virus and RSV were rare. Fourth, in the clinical symptoms and investigation association data, we reported some notable differences between virus-positive and -negative exacerbations that may assist in clinical decision making and future research.
In conclusion, although respiratory viruses (particularly HRVs and HPIVs) were frequently found during exacerbations of non-CF bronchiectasis, they were not as pervasive as has been described in cohorts of children with asthma. Codetection of viruses was also significantly less common than in paediatric asthma cohorts. Time-sequenced cohort studies during stable state, exacerbations and recovery periods are needed to determine the role of viral infections and their interaction, if any, with bacteria.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
- Data supplement 1 - Online supplement
Footnotes
-
Contributors NK monitored data collection for the whole study, performed the statistical analysis, interpreted the data and drafted the paper. He is the guarantor. IMM analysed airway samples for extended viral assay in the laboratory and revised the paper. TPS monitored the extended viral assay on airway samples in the laboratory. IBM monitored data collection, interpreted the data and revised the paper. ABC conceptualised the study, helped with the statistical analysis, helped with interpretation of data and revised the paper.
-
Funding Australian National Health and Medical Research Council (545216) (AC). ANZ Trustees PhD Scholarship supplemented with AC's grants (NK).
-
Competing interests None.
-
Ethics approval Queensland Health Children's Health Services Ethics Committee.
-
Provenance and peer review Not commissioned; externally peer reviewed.