In the last twenty years the term "asthma" has been applied to all wheezing illnesses in children replacing the terms "wheezy bronchitis" and "asthmatic bronchitis" which were used in the past to describe milder intermittent episodes of wheeze in young children ( 1,2). The change in nomenclature was based on studies that demonstrated that when children were diagnosed with "wheezy bronchitis", they were rarely treated with "asthma" medication even when they had very frequent, significant episodes of asthma ( 3,4 ). More recently the pendulum has begun to swing back. It has become apparent that there is heterogeneity in the respiratory conditions of childhood now lumped together as asthma ( 5-7 ). Children with symptoms of intermittent viral-induced asthma might have a different pathology than children with more chronic allergic asthma. However we do not have sufficient data on the pathophysiology of these conditions to be certain. Acute episode of asthma in children seem to commonly occur with viral infections whether the underlying asthma is chronic or intermittent in nature. It is proposed that intermittent viral-induced asthma in non-atopic children might lack significant eosinophilic inflammation in the airways compared to chronic allergic asthma. Confirmation of this hypothesis might lead to changes in predicted outcome and changes in recommended management.
Studies are now being published that suggest heterogeneity in childhood asthma and questioning the current asthma management guidelines especially in infants ( 6 ). Stevenson et al ( 7 ) studied children undergoing elective surgery by bronchoalveolar lavage and found that children with viral induced asthma, studied when they were free of infection, lacked eosinophils in the lavage compared to children with atopic asthma. We have previously published that pre-school children who had recurrent viral-induced wheeze but were non-atopic had significantly lower peripheral blood eosinophils and serum ECP than children with recurrent wheeze who were atopic ( 8 ). This data was among the first that suggested two different forms of asthma i.e. eosinophilic and pauci-eosinophilic. Since then epidemiologic data has suggested that the non-atopic children have a greater chance of "outgrowing" the asthma over time ( 9-11 ). We ( 8 ) and others have suggested that inhaled steroid treatment might not be necessary in young asthmatic children who are not atopic or in whom the atopy is not leading to significant eosinophilic inflammation ( 6,7 ). More recently we have found that children with persisting post-infectious cough lack eosinophils in their sputum, nasal passages and blood, yet have airway reactivity as indicated by positive methacholine challenges (Zimmerman, Silverman, Tarlo, Chapman, Kubay,) Certainly studies using serum ECP as a surrogate marker for eosinophilic inflammation in asthma suggest that even atopic asthmatics have varying levels of eosinophilic inflammation so that those children with low levels of eosinophilic activity might also have a better prognosis. All of this data suggests heterogeneity in childhood asthma and raises the question as to whether "non-eosinophilic" childhood asthma should be treated differently than "eosinophilic" asthma.
Intermittent recurrent episodes of cough and wheeze are felt clinically to represent the consequence of viral respiratory tract infections. These episodes are generally preceded by symptoms suggestive of infection including: fever; sore throat; and nasal discharge. The clinical attributes of these episodes have been described in two excellent prospective studies.
Clough and Holgate ( 12 ) described 600 episodes of respiratory symptoms associated with a defined fall in peak flow readings. Each child experienced an average of 3.5 episodes per year lasting a mean of 4.1 days (2 - 27 days). Atopy, including sensitivity to dust mite, had no effect on episode frequency, duration or severity. In this study no direct evidence of viral etiology was provided and a control group without history of "asthma" was not enrolled. These children did have positive methacholine challenges with atopic children having significantly more reactive airways ( 11 ).
Mertsola et al ( 13 ) followed 56 patients who had had recurrent wheezing episodes, prospectively for 3 months. They identified 115 episodes of respiratory tract symptoms. Virus or Mycoplasma was identified by culture in 52 (45%). The patients had a mean of 2.1 episodes lasting a total mean of 30 days of the 92 days studied. Wheezing occurred during 76 (66%) of the 115 episodes, a third of which required hospitalization because of severity. The incidence of wheezing was not associated with atopy, or positive virologic tests but was associated with parental smoking and more than one sibling. The study did not follow a control group and viral isolation occurred equally in episodes with and without wheezing.
There are three indirect lines of evidence that suggest an etiologic relationship between viral infection and exacerbation of asthma. These include: 1. clinical studies using a variety of culture and serologic techniques to identify viruses in acute episodes of asthma; 2. studies of the effect of viral infections on pulmonary function and 3. studies correlating the severity of viral illness with the degree of airway reactivity or symptoms.
The early data has been reviewed recently ( 14 ). The authors tabulated the results of cross sectional and prospective studies of virus identification in episodes of asthma. The 10 cross-sectional studies showed identification rates of 9.8 - 48.6% with an average of 24% of 4896 episodes. Eight prospective studies in children yielded 31.9% positive for any virus in 602 episodes. The rate for adults was much lower. Viral identification was always significantly greater during symptomatic compared to asymptomatic periods. In contrast bacterial isolation was no different between symptomatic and asymptomatic. The viruses identified included Rhinovirus, Coronavirus, Respiratory Syncytial Virus, Parainfluenza, Influenza and Adenovirus. It was suggested that the rate of identification would be higher when newer molecular technology would be applied to the problem.
More recently, the same authors have confirmed their prediction by publishing their experience in the identification of viruses in association with exacerbations of asthma in children aged 9 - 11 years. The viruses were identified by a combination of conventional methods and molecular methods i.e. polymerase chain reactions. It was possible to identify viruses in 80 - 85% of episodes with the most commonly identified virus by far ( i.e 75% of cases) being rhinovirus ( 15 ). Episodes were defined as a fall in peak flow readings and the median duration of these falls in peak flows were 14 days. This data essentially confirmed the clinical impression that acute episodes of cough and wheeze which occurred with sore throat, fever and nasal congestion were associated with viral illnesses especially rhinovirus.
Several studies in normal adults have documented changes in lung function consistent with airway obstruction after viral infection ( 16-19 ) and other studies have found that bronchial hyperresponsiveness to non-specific stimuli increase and persist following viral infection ( 20,21 ). However negative studies also exist. Halperin et al ( 22 ) exposed asthmatic volunteers to rhinovirus and although coryzal symptoms occured in 17, only 4 had 10% or greater decrease in FEV1 and an increase in histamine sensitivity. Jenkins and Breslin ( 23 ) studied the effect of upper respiratory tract infection on pulmonary function and Histamine challenge in a group of normal and asthmatic adult volunteers. There were no significant changes in any measurement during viral infection in either normal or asthmatic subjects. In contrast, Gern et al ( 24 ) reported a change in histamine PD 20 for subjects experimentally infected with rhinovirus who had allergic rhinitis compared to normal controls. The authors reported that host factors such as allergy, baseline FEV1 and baseline PD20 influenced the changes in lower respiratory physiology in response to rhinovirus infection. Grunberg et al.( 25 ) examined sputum induced in atopic, non-smoking, mildly asthmatic subjects experimentally infected with rhinovirus in a placebo controlled study. They found that experimental rhinovirus infection in these individuals increased airway hyperresponsiveness and increased airway inflammation with increase in eosinophils, ECP, IL-8 and IL-6 in the induced sputum.
Several possible mechanisms for viruses as precipitants of asthma symptoms have been proposed but none can be considered proven. It has been proposed that viruses may lead to exacerbations of asthma by any of several mechanisms including: 1. epithelial damage leading in turn to other effects such as increased neural sensitivity; 2. Increase in chemical mediators such as histamine, arachadonic acid metabolites, kinins; 3. Increase in cell driven inflammation through cytokines and cell activation; 4 Beta receptor down-regulation.
Lemanske et al (26 ) demonstrated that experimental rhinovirus infection increased the airway reactivity to specific allergen challenge i.e. ragweed sensitivity in ragweed allergic individuals and at the same time promoted late phase reactions to the allergen. This work was later extended by Calhoun et al (27) who demonstrated that when patients with allergic rhinitis were infected with rhinovirus 16 and then challenged with segmental airway installation of allergen, they demonstrated both increased airway reactivity to the allergen and increased influx of eosinophils compared to controls. Similarly Fraenkel et al (28) showed that experimental infection of atopic, mildly asthmatic volunteers with rhinovirus 16 resulted in increased histamine airway reactivity and on biopsy there was an increase in mucosal lymphocytes and eosinophils. This work was confirmed in the study already cited by Grunberg et al (25) where the induced sputum of atopic, mildly asthmatic adults infected with rhinovirus demonstrated increased eosinophils.
At first glance this body of evidence seems to conflict with the study of Stevenson et al (7) in children with viral induced asthma who underwent lavage when they were free of viral infection. These investigators did not find evidence for eosinophils in the airways. While it is possible that had those patients been studied during a viral induced episode of wheezing, they may have demonstrated eosinophils in the airway, it is just as likely that if those patients were non-atopic, they might not demonstrate eosinophils in the airway even when symptomatic. Clearly investigations need to be done on childhood asthmatics, both atopic and non-atopic when they are free of symptoms and when they are ill with an episode due to viral infection.