Prevalence and Impact of Respiratory Viral Infections in Young Children With Cystic Fibrosis: Prospective Cohort Study
OBJECTIVE. We aimed to investigate differences in upper and lower respiratory tract symptoms in relation to respiratory viral infections detected with polymerase chain reaction assays in young children with cystic fibrosis and healthy control subjects.
METHODS. In a 6-month winter period, 20 young children with cystic fibrosis and 18 age-matched, healthy, control subjects were contacted twice per week for detection of symptoms of an acute respiratory illness. If any symptom was present, then a home visit was made for physical examination and collection of nasopharyngeal swabs for viral analysis. In addition, parents were instructed to collect nasopharyngeal swabs every 2 weeks.
RESULTS. Children with cystic fibrosis and healthy control subjects had similar frequencies of acute respiratory illnesses (3.8 ± 1.0 and 4.2 ± 1.7 episodes, respectively). Although there were no significant differences in upper respiratory tract symptoms, the children with cystic fibrosis had longer periods of lower respiratory tract symptoms (22.4 ± 22.2 vs 12.8 ± 13.8 days) and a higher mean severity score per episode (2.35 ± 0.64 vs 1.92 ± 0.46). In addition, similar increases in upper respiratory tract symptom scores were associated with significantly greater increases in lower respiratory tract symptom scores in children with cystic fibrosis. No differences in the seasonal occurrences and distributions of respiratory viruses were observed, with picornaviruses and coronaviruses being the most prevalent.
CONCLUSIONS. Although there were no differences in the seasonal occurrences and distributions of polymerase chain reaction-detected respiratory viruses, acute respiratory illnesses were frequently associated with increased lower respiratory tract morbidity in young children with cystic fibrosis.
Viral infections are the most common cause of acute respiratory symptoms in otherwise healthy children. The relationship between viral infections and chronic respiratory diseases such as asthma has long been recognized. It has been estimated that up to 85% of asthma attacks in children in the community are associated with respiratory viral infections.1 The role of respiratory viral infections for patients with cystic fibrosis (CF) is less clear, however.
A few prospective studies suggested that the clinical impact of respiratory viral infections on patients with CF is more severe than the virus-related morbidity in healthy control subjects. Viral respiratory tract infections in children and adults with CF seem to be associated with significant risks of pulmonary exacerbation, hospitalization, and decreased lung function.2–6 In addition, respiratory viral infections seem to predispose patients to secondary bacterial colonization and infection.7,8 The differences in the clinical impact of respiratory viral infections between patients with CF and healthy control subjects might be attributable to differences in the frequency or distribution of viruses or to patient factors related to CF. The few studies to date have yielded no evidence that viral infections are more common in individuals with CF.3,6,9
Several methodologic characteristics of those studies need to be addressed, however. Most studies reported data only for hospitalized patients, lacked the use of healthy control subjects, or failed to present longitudinal data. In addition, earlier reports might have underestimated the prevalence of viral infections because relatively insensitive viral detection methods were used, such as serological tests and tissue cultures.2,5,6,9 More recently, polymerase chain reaction (PCR) techniques have improved the detection of respiratory viruses.10 Finally, only sparse data on the role of newly detected viruses implicated in lower respiratory tract infections in young children (eg, human metapneumovirus and coronavirus NL63) are available in relation to CF. This study was designed to investigate differences in the prevalence and clinical impact of virus-associated acute respiratory illnesses (ARIs) between young children with CF and healthy control subjects, by using sensitive PCR detection methods.
Study Design and Subjects
A prospective longitudinal study was conducted during a 6-month respiratory virus season, from November until May. A total of 20 young children with CF (age: 0–7 years) were enrolled and completed the study. Diagnosis of CF was based on the presence of ≥2 of the following criteria: sweat chloride level of >60 mEq/L, positive genetic test results, and clinical features consistent with CF. All children had tested negative for Pseudomonas aeruginosa until the start of the study. A total of 19 age-matched, healthy, control subjects were enrolled; 1 failed to complete the study. All had a negative history of respiratory diseases and other major diseases. The study was approved by the local medical ethics review committee (University Medical Centre Utrecht), and the parents gave written informed consent.
At enrollment, demographic data, the child's medical history, and environmental risk factors were documented by using questionnaires. Physical examinations were performed and throat or cough cultures were obtained to determine bacterial colonization. Parents were instructed to obtain nasopharyngeal swab specimens, for viral PCR testing, from their child every 2 weeks. The swabs were sent by mail to the study coordinator within 24 hours. Throughout the study period, 1 of 2 physicians acting as study coordinators (Drs van Ewijk and van der Zalm) contacted the families twice per week by telephone or e-mail. Any symptom of a respiratory illness was assessed with a standard questionnaire and directly tabulated. When an ARI was identified, the study coordinator made a home visit within 48 hours, to perform a physical examination and to obtain a nasopharyngeal swab for viral PCR testing. Patients with CF were treated according to the guidelines of the CF Centre Utrecht. Antibiotics were prescribed by the treating pediatrician or family doctor. No specific antiviral treatment or prophylaxis was prescribed for patients with CF or healthy control subjects.
Respiratory Illness Questionnaire
The twice-weekly respiratory illness questionnaire concerned both upper respiratory tract symptoms (URTSs) and lower respiratory tract symptoms (LRTSs), as defined in previous studies.3,9 Recorded parent-reported URTSs were coryza (rhinorrhea or nasal congestion), sore throat, ear ache, ear discharge, and temperature of >38°C. A score of 0 was given if a symptom was absent and 1 if it was present, yielding a maximal score of 5 points. Recorded parent-reported LRTSs were increased cough, increased sputum production or productive cough, shortness of breath, reduced exercise tolerance, decreased appetite, and malaise. On the basis of the patient diary information, a score of 0 was given if a symptom was absent, a score of 1 if there was a moderate increase, compared with baseline, and a score of 2 if the symptom was severe, giving a maximal score of 12. An ARI was diagnosed if the child had coryza or one of the other URTSs in combination with a temperature of >38°C or had a total LRTS score of ≥2. For diagnosis of a new ARI episode, the total symptom score had to have returned to 0 for ≥1 week.
Physical examinations were performed at the beginning and end of the study (baseline if no respiratory symptoms) and at each home visit (with respiratory symptoms) after diagnosis of an ARI. Examinations included standardized measurements of weight, respiratory rate, and transcutaneous oxygen saturation. Assessment of retractions, ear-throat-nose examination, and auscultation of the lungs were performed. Signs of rhinitis, pharyngitis, otitis, and abnormal auscultation findings (rales or crackles) were assessed.
Respiratory viral infection was established through detection of virus from nasopharyngeal swabs by using standard PCR techniques, as described previously.11 Specimens were collected with cotton-tipped swabs from both the nose and the posterior oropharynx. Both swabs were inserted into a single vial containing gelatine lactalbumine yeast virus transport medium with pimaricine (0.1 mg/mL) as viral transport medium. Samples were stored at −20°C until analysis. PCR assays were performed at the National Institute of Public Health and the Environment (Bilthoven, Netherlands). Both nasopharyngeal swabs were tested for rhinovirus, enterovirus, coronavirus 229E, OC43, and NL63, human metapneumovirus, respiratory syncytial virus A and B, influenzavirus A and B, and adenovirus, as well as Mycoplasmapneumoniae and Chlamydia pneumonia, by using standard PCR techniques.
Comparisons of the distributions of categorical variables between groups were examined by using the χ2 test or Fisher's exact test and the means of continuous variables by using the 2-tailed Student's t test of independent variables or the nonparametric Mann-Whitney U test. The correlation between the total (seasonal) URTS score and total (seasonal) LRTS score for each individual for patients with CF and healthy control subjects was assessed with Pearson's standard regression analysis. Multivariate linear regression analysis was used to analyze differences in increments for total LRTSs per total URTSs between patients with CF and healthy control subjects. A significance level of P < .05 was used throughout. The analyses were performed by using SPSS 12.0 (SPSS, Chicago, IL).
There were no significant differences between the children with CF and the healthy control subjects with respect to age, weight, respiratory rate, number of siblings, or day care or school attendance at baseline (Table 1). The children with CF had lower baseline pulse oxygen saturation values (P < .01), 65% had been vaccinated for influenza, and 70% used antibiotic prophylaxis throughout the study period, compared with none of the healthy control subjects (both P < .001). Children with CF more often had positive bacterial throat or cough swab culture results at the start of the study, compared with healthy control subjects (P < .05), mainly for Staphylococcusaureus and Haemophilus influenzae.
Data for the 6-month period were plotted for each child, to allow visual inspection of respiratory illness patterns (Fig 1). The coded data were analyzed. On the basis of data from patient diary records, control subjects and children with CF had similar numbers of ARIs, resulting in similar numbers of home visits. No differences in the mean URTS score per episode were found (Table 2). Children with CF tended to have longer periods of URTSs; however, the difference did not reach statistical significance (P = .1). For LRTSs, there were significant differences between children with CF and healthy control subjects in both the duration and severity of the episodes. Children with CF had longer periods of LRTSs (P < .01) and had higher mean LRTS scores per episode (P < .05). Physical examinations during ARI periods showed a significantly greater increase in respiratory rate (P < .001) and a greater decrease in pulse oxygen saturation (P < .01) for children with CF, both compared with baseline measurements. In addition, retractions (P < .001) and abnormal auscultation findings (P < .01) were observed more frequently for children with CF. Antibiotic courses were prescribed by a family physician or treating specialist 24 times for 13 children with CF, compared with 2 courses for 2 healthy control subjects (P < .01).
No differences in separate URTS score items were found between patients with CF and healthy control subjects (Table 3). Temperature of >38°C was reported more frequently for healthy control subjects, whereas the LRTS score items of increased cough (P < .05), increased sputum or productive cough (P < .01), and shortness of breath (P < .05) were reported more often for children with CF. To study the relationship between URTSs and LRTSs, the sum of all URTS scores was plotted against the sum of all LRTS scores for the whole period for each individual (Fig 2). High URTS scores were associated with high LRTS scores for each group (R2 = 0.40 for healthy control subjects and R2 = 0.55 for children with CF). For children with CF, an increase in URTS scores was associated with a significantly greater increase in LRTS scores, compared with healthy control subjects (P < .01).
Detection of Respiratory Pathogens
PCR analyses for viral and atypical pathogens were performed with comparable numbers of specimens in the 2 groups, with mean values of 17.6 ± 2.64 samples per patient for the CF group and 16.3 ± 2.16 samples per patient for the control group (P = .1). Table 4 presents the seasonal occurrence of the different pathogens during the study period. No significant differences were found between the 2 groups. All children in both groups had an infection with rhinovirus during the study period, followed by enterovirus and coronavirus as the most prevalent viruses.
Our study showed that ARIs were associated with increased LRTSs for children with CF, compared with healthy control subjects. Although there were no differences in the frequency of ARIs or mean URTS scores, children with CF had longer and more-severe periods of LRTSs. Similar increases in URTS scores were associated with significantly greater increases in LRTS scores for children with CF. No differences in the occurrence or distribution of respiratory viruses were observed between the 2 groups. These findings argue that patient factors related to CF play an important role in the differences in clinical courses.
Our findings emphasize that CF is a disease that involves primarily the lower airways. During ARIs, patients with CF more frequently present LRTSs as cough, sputum production, and shortness of breath. In our study, physical examinations confirmed the reported symptoms. In children with CF, ARIs caused a significantly greater increase in respiratory rate, a greater decrease in pulse oxygen saturation, and more retractions and abnormal auscultation findings, compared with control subjects. Earlier data from different study settings showed that URTSs were associated with LRTSs in 31% to 76% of patients with CF.2,12 Upper respiratory tract infections were related to decreases in lung function,2,3,5,6 greater use of antibiotics,5 and high hospitalization rates.3,6,12,13 Surprisingly, in our study none of the children in either group was hospitalized during the study period. This might be attributable to the young age of the children; probably most of them had not developed important structural lung abnormalities before this study. Also, the extensive use of antibiotic prophylaxis and the low threshold for prescription of antibiotic courses for the children with CF in our center might have prevented hospital admissions.
In this study, we used URTS and LRTS scores, as described by others.3,6,9,14,15 A lack of validated symptom scores for young children with CF might result in reporter and observer bias. In our study, however, symptoms reported by the parents, clinical findings of the investigators, and objective parameters (respiratory rate and oxygen saturation) all pointed in the same direction. This argues for reasonable internal validity of our symptom scores. In our study population, the diagnosis of ARI was never made on the basis of nonrespiratory symptoms only.
Several patient factors in CF might be suggested to cause the differences in the clinical impact of viral infections on LRTSs between children with CF and healthy control subjects. Firstly, children with CF were more frequently colonized with potential pathogenic bacteria at enrollment. The higher prevalence of S aureus in children with CF is in line with expectations and is known to cause pulmonary morbidity.16 Apparently, the relatively high rate of antibiotic prophylaxis targeting S aureus in the children with CF could not prevent these differences. Secondly, the greater impact of a viral infection on LRTSs might be the manifestation of a diminished specific antiviral defense. Zheng et al17 found greater viral replication in CF airway epithelium and increased production of proinflammatory cytokines because of an impaired nitric oxide synthase 2 signaling pathway, which is an important part of the innate antiviral defense. More recently, Colasurdo et al18 showed that mice with CF have an exaggerated inflammatory response but impaired ability to clear respiratory syncytial virus. It might be speculated that this greater viral replication and exaggerated inflammation cause more local damage in the airways. In our study, virus-associated ARIs occurred equally in children with CF and healthy control subjects; however, we did not quantify viral loads, which might have supported these findings. Furthermore, it might be very interesting to investigate the different influences of individual viruses on the level of LRTSs in patients with CF, compared with healthy control subjects, in a larger study. Because of the relatively small numbers of patients, our data lack sufficient power for such an analysis.
Thirdly, synergism between bacteria and viruses might contribute to the greater impact of ARIs. Several studies suggest that respiratory viral infections facilitate P aeruginosa acquisition and colonization.2,7,8,13,19 In patients with intermittent or chronic P aeruginosa infection, a viral infection is often followed by an increase in antipseudomonal antibody levels.8 It might also be speculated that structural abnormalities of the lower airways in patients with CF are responsible for the increased lower respiratory tract morbidity. In contrast to earlier studies,4,6,9,12,20 in our study we focused on young children with few structural abnormalities in their lower airways. In older patients with more-advanced lower airway abnormalities, the differences may be even more pronounced.
The differences in clinical impact do not seem to be related to virus-specific characteristics. The seasonal frequencies and distributions of the different viruses and atypical pathogens, as presented, were similar for children with CF and healthy control subjects. Our results are comparable to earlier findings for children in the general population.21 Equal numbers of viral infections were observed for patients with CF and healthy control subjects,6,9 and 1 study described an increased number of viral infections in healthy control subjects.3 However, earlier studies might have underestimated the prevalence of viral infections because they used relatively insensitive detection methods, such as serological assays and tissue cultures.2,5,6,12 We used more-sensitive PCR methods to detect viruses and atypical pathogens. Only a few CF studies have used PCR methods for this purpose. Two studies used PCR methods for the detection of picornaviruses; they reported that 43% of samples tested positive for picornaviruses in individuals with colds2 and 16% tested positive for rhinovirus in cases of pulmonary exacerbation.5 A third study used PCR methods for detection of several viral and atypical pathogens and found a low 16% rate of positive nasopharyngeal aspirates at regular outpatient visits, irrespective of symptoms and without comparison with healthy control subjects.22 To the best of our knowledge, our data are the first results concerning the newly detected coronavirus NL63 in relation to CF, with no differences between children with CF and healthy control subjects. This virus is known to be implicated in lower respiratory tract infections in young children in general.23,24
Our findings emphasize the importance of viral respiratory infections in lower respiratory tract morbidity in young children with CF. Modern treatment strategies for pulmonary complaints are mainly aimed at treating bacterial infections with antibiotics and sputum evacuation. With improving survival rates for young patients with CF because of better antibacterial treatment, antiviral therapy might become a new therapeutic goal in the future. Increasing evidence that viral respiratory infections, possibly in synergism with bacteria, play an important role in irreversible local damage emphasizes the possible importance of prevention of viral infections (for example, through active or passive immunization). However, influenza vaccination could not prevent an influenza infection in 1 patient with CF in our study. Viral inhibitors used in an early phase of a viral infection, antibiotic treatment or prophylaxis during a viral infection, and development of specific virus-bacterium interaction blockers might be interesting options.25
This study clearly shows that ARIs are associated with increased LRTSs in young children with CF, compared with healthy children. These first results of extensive viral analysis with PCR methods in CF demonstrate no major differences in seasonal frequencies or distributions of respiratory viral infections. Patient factors related to CF seem to play an important role in these differences in clinical courses. Our findings confirm the suggestion that viral respiratory infections play an important role in pulmonary morbidity in young children with CF, and they might suggest new therapeutic strategies to improve the prognosis for patients with CF.
This work was supported by Wilhelmina Children's Hospital Research Fund (fellowship to Dr van Ewijk).
We thank Toyba Yimam and Bianca van der Zwan from the Laboratory for Infectious Diseases and Screening, National Institute of Public Health and the Environment, for assistance in performing the PCR studies.
- Accepted March 10, 2008.
- Address correspondence to Bart E. van Ewijk, MD, Cystic Fibrosis Centre and Department of Pediatric Respiratory Medicine, Wilhelmina Children's Hospital, University Medical Centre Utrecht, PO Box 85090, Office KH 01.419.0, 3508 AB Utrecht, Netherlands. E-mail:
Financial Disclosure: Dr van der Ent received unrestricted research grants from Roche, Grunenthal, and GlaxoSmithKline in 2005 and 2006.
What's Known on This Subject
Respiratory viral infections may have a large impact on short- and long-term pulmonary morbidity in cystic fibrosis. Differences in the prevalence and clinical impact of virus-associated acute respiratory illnesses between patients with cystic fibrosis and healthy control subjects are unclear.
What This Study Adds
This community-based, prospective study shows, using sensitive PCR detection methods, that patients with cystic fibrosis have longer and more-severe periods of virus-associated, lower respiratory tract symptoms, compared with healthy control subjects, without differences in seasonal viral occurrences and distributions.
- ↵Johnston SL, Pattemore PK, Sanderson G, et al. Community study of role of viral infections in exacerbations of asthma in 9–11 year old children. BMJ.1995;310 (6989):1225– 1229
- ↵Collinson J, Nicholson KG, Cancio E, et al. Effects of upper respiratory tract infections in patients with cystic fibrosis. Thorax.1996;51 (11):1115– 1122
- ↵Hiatt PW, Grace SC, Kozinetz CA, et al. Effects of viral lower respiratory tract infection on lung function in infants with cystic fibrosis. Pediatrics.1999;103 (3):619– 626
- ↵Ong EL, Ellis ME, Webb AK, et al. Infective respiratory exacerbations in young adults with cystic fibrosis: role of viruses and atypical microorganisms. Thorax.1989;44 (9):739– 742
- ↵Smyth AR, Smyth RL, Tong CY, Hart CA, Heaf DP. Effect of respiratory virus infections including rhinovirus on clinical status in cystic fibrosis. Arch Dis Child.1995;73 (2):117– 120
- ↵Johansen HK, Hoiby N. Seasonal onset of initial colonisation and chronic infection with Pseudomonas aeruginosa in patients with cystic fibrosis in Denmark. Thorax.1992;47 (2):109– 111
- ↵van Gageldonk-Lafeber AB, Heijnen ML, Bartelds AI, Peters MF, van der Plas SM, Wilbrink B. A case-control study of acute respiratory tract infection in general practice patients in the Netherlands. Clin Infect Dis.2005;41 (4):490– 497
- ↵Goss CH, Burns JL. Exacerbations in cystic fibrosis, part 1: epidemiology and pathogenesis. Thorax.2007;62 (4):360– 367
- ↵Zheng S, Xu W, Bose S, Banerjee AK, Haque SJ, Erzurum SC. Impaired nitric oxide synthase-2 signaling pathway in cystic fibrosis airway epithelium. Am J Physiol Lung Cell Mol Physiol.2004;287 (2):L374– L381
- ↵van Ewijk BE, Wolfs TF, Fleer A, Kimpen JL, van der Ent CK. High Pseudomonas aeruginosa acquisition rate in CF. Thorax.2006;61 (7):641– 642
- ↵Kahn JS. Epidemiology of human metapneumovirus. Clin Microbiol Rev.2006;19 (3):546– 557
- ↵van der Hoek L, Pyrc K, Berkhout B. Human coronavirus NL63, a new respiratory virus. FEMS Microbiol Rev.2006;30 (5):760– 773
- Copyright © 2008 by the American Academy of Pediatrics