Relationship between respiratory syncytial virus bronchiolitis and future obstructive airway diseases

Respiratory syncytial virus infections

RSV has been described as the single most important virus causing acute respiratory tract infections in infants and toddlers. RSV infections have a seasonal pattern, with annual epidemics during winter and spring in temperate climates, and in the rainy season when the temperature falls in tropical climates. The peak infection rates occur in infants of 6 weeks to 6 months of age. By the age of 2 yrs, ∼90% of young children have been infected with RSV 4, 5.

The virus

RSV was first identified in 1956 as the agent that caused chimpanzee coryza 6. The virus belongs to the Paramyxoviridae family, where it has been assigned a genus of its own, Pneumovirus. RSV is a large virion which consists of a negative RNA genome in a helical nucleocapsid surrounded by an envelope (fig. 1⇓). The envelope contains a viral attachment glycoprotein (G) and a fusion glycoprotein (F). These glycoproteins are the major surface antigenic determinants of RSV. The F and G proteins are vital for the infectivity and pathogenicity of RSV. The F protein is relatively stable, which makes it a suitable target for therapeutic intervention. The G protein is more variable and forms the basis for subgroup classification of RSV into type A and type B. It has been proposed that RSV type A should be the most pathogenic, although current opinion suggests that there are no certain clinical differences between the two subgroups 4, 5.

Paramyxoviridae are transmitted in respiratory droplets and initiate infection in the respiratory tract. The virions penetrate the cell by fusion with the plasma membrane, and the virus replicates in the cytoplasm. Viruses induce cell-to-cell fusion, thereby causing multinucleated giant cells to form. These large fused cells, syncytia, gave RSV its name.

Clinical features of respiratory syncytial virus bronchiolitis

RSV bronchiolitis typically affects the very young infant, often only 1–3 months old, mostly <6 months of age. RSV infects via the upper respiratory tract, particularly through the nasopharynx, and the eyes. The incubation period is 3–5 days. Infection of the bronchiolar epithelial cells results in mucosal inflammation and oedema. Necrosis of epithelial cells and intraluminal plugs, consisting of mucus and cellular debris, cause a ball-valve airway obstruction which leads to hyperinflation of the distal airways and alveoli 4, 5.

The first symptoms of RSV bronchiolitis are coryza and dry cough followed by increasing breathlessness. Sometimes apnoeic spells are observed in the initial stage of the disease in infants <2 months of age, especially in babies born preterm. If fever is present, it is usually low grade. The baby feeds poorly. Hypoxaemia, which often leads to cyanosis, follows, and carbon dioxide retention may lead to hypercarbia. Other clinical symptoms are a sharp, dry cough, tachypnoea and tachycardia. Subcostal and intercostal retractions are seen. The chest becomes hyperinflated due to small airways obstruction and air trapping. Chest radiographs show over-inflated lungs and often perihilar infiltrates. On chest auscultation, fine end-inspiratory crepitations are heard, and the expiratory phase is prolonged. Wheezes may be audible with or without a stethoscope 4.

Diagnosis

RSV is the most common agent, which can be detected in infants with bronchiolitis. The virus can be demonstrated in a positive immunofluorescence antibody or an enzyme-linked immunosorbent assay (ELISA) test performed on nasopharyngeal aspirates or washes, or by cell culture 8. Overall, RSV has been detected in 40–90% of bronchiolitis cases 9–12. During the peak of an RSV outbreak, ≥80% of bronchiolitis cases are associated with RSV 5.

Treatment

The treatment of RSV bronchiolitis is mainly supportive and symptomatic 4. This includes gentle handling, moderate fluid supply and maintaining oxygenation at arterial oxygen (O₂) saturation (S_a,O2) ≥95% by administration of humidified O₂. The efficacy of β₂-agonists in bronchiolitis has been questioned by studies showing no significant effect 13. A meta-analysis of eight randomized controlled studies, including 324 children <1 yr old, indicates a moderate, short-lasting improvement in symptom score and S_a,O2 by inhaled salbutamol (0.1–0.15 mg·kg⁻¹ body weight) 14. Some studies indicate that the effect of inhaled adrenaline or racemic adrenaline is better than that of selective β₂-agonists, possibly due to the α-agonist action of adrenaline. In a double-blind randomized controlled trial (RCT) study, Kristjánsson et al. 15 demonstrated a small, but significant, improvement of oxygenation and symptom score by inhaled racemic adrenaline. In fact, there are four double-blind, RCT studies that report a better effect of inhaled adrenaline or racemic adrenaline, compared to salbutamol, in the treatment of acute bronchiolitis in infants 16–19. In contrast, almost all placebo-controlled studies on the use of corticosteroids, given by systemic administration or inhalation, show no improvement in the clinical course of acute bronchiolitis 20–24. van Woensel et al. 25 found that oral prednisolone seemed to accelerate the clinical recovery of children admitted to hospital with RSV bronchiolitis, but the results of that study are an exception. Taken together, the studies indicate that there is no significant effect of corticosteroids in the acute phase of RSV bronchiolitis. The results of antiviral treatment with aerosolized ribavirin are also disappointing. The therapeutic efficacy in the acute infection has been questioned, and in a meta-analysis, no statistically significant effect was demonstrated 26.

Neuroimmune interactions

Experiments using animal models demonstrate that RSV causes acute and chronic changes in neural control of the airway, altering airway control, resulting in decreased relaxant and increased contractile responses in airway smooth muscle. When infection occurs in early life the alterations persist for long periods 31.

Recent studies indicate that immune and neural mechanisms may be linked and that post-RSV airway inflammation may partly be explained on the basis of such neuroimmune interactions 32. In lungs from RSV-infected rats, Piedimonte et al. 33 have demonstrated an increased vascular permeability elicited by capsaisin stimulation of unmyelinated sensory nerves. The increase in vascular permeability resulted from upregulation of the high-affinity receptor for substance P (the neurokinin (NK)-1 receptor). Such exaggerated neurogenic inflammatory responses can be observed long after RSV has been cleared from the lungs, which suggests that the underlying mechanisms may become independent of the presence of the virus. It is suggested that after resolution of the acute RSV infection, stimulation of the sensory nerves by an airborne irritant may induce inflammation via NK-1-expressing T-lymphocytes, which retain a nonspecific memory of the early infectious episode 32.

Chronic persistent respiratory syncytial virus infection

RSV may persist in the infected cells after the initial infection and it may constitute a foreign protein, leading to a chronic inflammation 34, 35. Guinea pigs, experimentally inoculated with human RSV, show histological evidence of acute bronchiolitis and chronic persistence of viral antigens and viral genome in the lungs 36. Infected animals develop an anti-RSV immunoglobulin (Ig)-G1 antibody response (the main class of antibody involved in guinea pig allergic responses), histological evidence of acute bronchiolitis, and chronic airway inflammation 36. Riedel et al. 37 have demonstrated that in guinea pigs, RSV infection of the airways causes persistent airway hyperresponsiveness (AHR) over a period of ≥5 weeks. During this period, viral antigen remained detectable in the lungs and may be responsible for ongoing AHR 37. From these studies, it appears that persistent RSV lung infection may be important in the pathogenesis of postbronchiolitis wheezing and asthma in children.

Surfactant dysfunction

Regarding nonimmunological factors, van Schaik et al. 38 have suggested that inadequate pulmonary surfactant function may be an important factor in the pathophysiology of RSV-induced bronchiolitis. In experiments with RSV-infected mice, the degree of surfactant dysfunction correlated with the presence of inflammatory cells in bronchoalveolar lavage fluid (BALF) 39. Surfactant abnormalities have also been demonstrated in infants with severe viral brochiolitis 40.

Postbronchiolitis effects on hyperresponsiveness and lung function

It is well known that viral respiratory infections transiently increase bronchial reactivity 69. This was observed in normal subjects after upper respiratory infections >35 yrs ago 70.

Several long-term follow-up studies of bronchial reactivity, after hospitalization with proven RSV bronchiolitis, have reported bronchial hyperresponsiveness to exercise and histamine several years later, even at school age 48, 49, 53, 71. However, in mild cases there seemed to be no hyperresponsiveness present at 8–12 yrs follow-up 72, 73.

Most follow-up studies of RSV bronchiolitis in infancy show that forced expiratory flow rates (e.g. forced expiratory volume in one second (FEV₁), are lower at school age compared with control groups 74. Kattan et al. 72 demonstrated early on that children who had been hospitalized for RSV bronchiolitis as infants, later had lung function abnormalities similar to those found in children with asthma. Decreased expiratory flow rates in children with a history of bronchiolitis, compared with controls, have also been reported in other follow-up studies 48, 49.

Respiratory syncytial virus bronchiolitis and subsequent wheezing

As long as 40 yrs ago, Wittig and Glaser 75 and Eisen and Bacal 76 reported that bronchiolitis in infancy was often followed by recurrent episodes of wheezing. In 1971, Rooney and Williams 77 reported that 56% of children who had been hospitalized with RSV bronchiolitis as infants had multiple wheezing episodes 2–7 yrs later. A family history of asthma was present in 72% of the patients who had recurrent wheezing episodes, compared with 18% of patients where no subsequent wheezing had occurred. After that study some 30 yrs ago, several controlled follow-up studies were performed. In a 10-yr follow-up, Pullan and Hey 49 reported that 42% of children with a history of RSV bronchiolitis in infancy had had further episodes of wheezing, while only 19% of controls had ever wheezed. The difference between RSV and control groups was most pronounced during the first 4 yrs of life, and many children with recurrent wheezing during that time had stopped wheezing by the time they were 6 yrs old. At 10 yrs, asthma was diagnosed in 6.2% of the bronchiolitis group versus 4.5% in the control group. Similar results were presented by Mok and Simpson 50 in a 7-yr follow-up; however, the outcome of RSV-induced bronchiolitis, pneumonia or bronchitis in infancy was not separated in that study. McConnochie and Roghmann 78 studied cases of mild bronchiolitis in infancy not requiring hospitalization, and they found a significantly increased risk of wheezing at 8 yrs of age. However, the relative risk of wheezing decreased during the following years, and by the time the children were 13 yrs old the relative risk of asthma was no longer significantly increased compared with the controls 78. A study that has been particularly valuable to shed light on the natural history of childhood wheezing is the longitudinal Tucson study 3. In that study, Stein et al. 3 reported that early lower airway infection with RSV is an independent risk factor for recurrent wheezing up to the age of 11 yrs, but not at 13 yrs. Thus, children with a history of bronchiolitis tend to have recurrent episodes of wheezing or asthma although episodes of wheezing tend to diminish by adolescence 79. Nonetheless, it is important to realize that otherwise healthy children treated for an early RSV infection constitute only a minority, at most 10%, of the children who will be treated for obstructive airways disease later on.

In the Swedish study by Sigurs et al. 2, asthma was significantly more common by the age of 7 yrs among the group of children who had been hospitalized with severe RSV bronchiolitis as young infants, 23% having current asthma versus 2% in the control group (fig. 2⇓). Corresponding figures for cumulative asthma were 30% versus 3%.

Sigurs et al. 2 also report that the development of asthma is often seen in the group of children who have had bronchiolitis, but who lack heredity for asthma. The conclusion of the Swedish investigators is that the early RSV bronchiolitis has induced a process which has led to asthma.

Kneyber et al. 80 recently published a meta-analysis from six follow-up studies of RSV bronchiolitis published between 1978–1998 45, 48–50, 53, 81 (fig. 3⇓). For all of these studies, postnatal age of initial illness was <12 months, all children were hospitalized for RSV bronchiolitis in infancy, the diagnosis RSV was virologically confirmed, and a control group was used. The meta-analysis confirms that wheezing is common after RSV bronchiolitis in infancy and may persist for ≥5 yrs of follow-up. Up to 5 yrs of follow-up after the RSV bronchiolitis, 40% of children reported wheezing, compared with only 11% in the control group (p<0.001). Between 5–10 yrs of follow-up, 22% of the bronchiolitis group reported wheezing, compared with 10% of the control group (p=0.19). The incidence of recurrent wheezing as defined by ≥3 wheezing episodes, also decreased with increasing years of follow-up, and ≥5 yrs of follow-up the difference between RSV and control groups was no longer significant. Regarding personal history of atopy, a family history of atopy and/or asthma, no significant differences between the RSV bronchiolitis and the control group were found. Kneyber et al. 80, therefore, concluded that it was unlikely that RSV bronchiolitis is a cause of atopic asthma later in life. Table 1⇓ summarizes the studies on RSV bronchiolitis and the occurrence of subsequent wheezing.

To conclude, several prospective case-control studies of high quality show that RSV bronchiolitis is often associated both with recurrent wheezing and asthma during a period of several years after the illness. However, wheezing tends to diminish, and most studies show no significant increase in wheezing by school age or adolescence compared with controls.

Does wheezing occur exclusively subsequent to respiratory syncytial virus bronchiolitis?

The association between virus infections and wheezing is also obvious later on during early childhood 8, 82, 83. Eriksson et al. 84 pointed out that, regardless of whether a group was recruited according to confirmed aetiology, such as RSV 45, 85, 86, or according to clinical symptoms (lower respiratory tract infection) 53, 83, 87, follow-up studies showed that at least 50% of the children studied had recurrent wheezing episodes. The authors, therefore, addressed the question of whether lower respiratory tract infection with RSV was more likely to induce later wheezing than other viruses 84. Hence, they examined the risk of subsequent wheezing in young children hospitalized for influenza A or RSV infection during a season with outbreaks of RSV and influenza A. Children with RSV were younger than those with influenza A (mean age 2.2 versus 11 months); otherwise, there was no major difference between the RSV and influenza A groups with respect to patient characteristics or environmental risk factors. The authors found no influence on the risk of later wheezing from type of viral infection. Sixty per cent of children had two or more episodes of wheezing after either influenza A or RSV. Obviously, the ages of the studied groups differed and the study does not say anything about similarities or differences in the mechanisms behind the subsequent wheezing. However, the study does illustrate that a tendency to subsequent wheezing is not a phenomenon exclusive for severe RSV infection.

A Swedish longitudinal study included all children admitted to hospital due to wheezing bronchitis before the age of 2 yrs (30% with RSV) 83. Thirty per cent of the children had persistent asthma at the age of 10 yrs 88. Children, in whom RSV was detected at admission to the study, were not over represented as compared with children in whom other viral agents were detected at first admission. Instead, persistent asthma correlated significantly with the recent presence of other atopic diseases in the subjects. Similar results have been presented by Korppi and co-workers 87, 89. The findings of these studies fit better with the view that asthma developed in predisposed children rather than with the hypothesis that RSV infection induced a process leading to persistent asthma. In conclusion, lower respiratory tract infections in young children, including those elicited by viral agents other than RSV, are often followed by repeated wheezing episodes.

Respiratory syncytial virus bronchiolitis and subsequent allergic sensitization

Development of asthma and development of allergies are not identical. As mentioned previously, it has not been clearly established whether the increased risk of subsequent wheezing after RSV bronchiolitis is linked to an increased risk of allergic sensitization. Most studies do not report an increased risk of allergic sensitization in children with a history of RSV bronchiolitis (table 2⇓). An increased risk of allergic sensitization was not found in the Tucson study 3. In the study group followed by Sigurs et al. 2, an increased allergic sensitization seemed to be evident both from positive skin-prick tests and from “any positive allergy test” (fig. 2⇓). By the age of 7 yrs, current atopic asthma was found in 8.5% of cases versus 1% in the control group. Allergic rhinoconjunctivitis was found in 14.9% of cases versus 2% of the controls, while the prevalence of atopic dermatitis was similar in the two groups. In addition to Sigurs et al. 2, Murray et al. 53 reported an increased risk of skin-prick test sensitization at the age of 6 yrs. However, when the same children were re-investigated at 10 yrs of age, an increased risk was no longer found 52 (table 2⇓).

As can be concluded from the summary of studies presented in table 2⇓, the evidence for an increased risk of allergic sensitization is not nearly as strong as the evidence for an increased risk of subsequent wheezing. In fact, most studies do not show a significant increase in personal atopy. Thus, it can be concluded that the increased risk of subsequent wheezing after RSV is not linked to an increased risk of atopy.

Possible reason for the difference in outcomes

A possible reason for diverging outcomes in different follow-up studies of early RSV infection is varying severity of the initial infection. Some 80–90% of children are infected with RSV during infancy 4, 5. Most have a subclinical or mild upper airway infection. Only a fraction of the children have lower respiratory airway infection; the prevalence has been estimated at 10–40%. Some 5% develop more severe lower airway symptoms. About 1–2% are hospitalized, and only a minority of the hospitalized children require intensive care. Such differences are likely to influence the outcome of the infection. Thus, the results from studies of children hospitalized for severe RSV bronchiolitis will probably differ from children seen as outpatients for wheezing, and most certainly from those only reported by parents as having had wheezing in the first year of life 90. It is likely that it is in severe early RSV infections that the virus is able to alter the response of the host to the current infection as well as to subsequent infections 2, 45. The difference in results between, for example, the Tuscon study by Stein et al. 3 and the Swedish studies by Sigurs and co-workers 2, 45, may have occurred because the Tuscon study seems to be based on relatively mild cases of bronchiolitis, while the Swedish studies concerned severely ill infants needing hospitalization.

Causation or association?

As pointed out by Long et al. 85 and Sigurs et al. 2, descriptive studies can suggest connections between RSV and subsequent symptoms. McBride 90 states that since most studies have been observational, the fundamental question of causation versus association remains unresolved. Does severe RSV infection during infancy cause the differences in pulmonary function observed later in life, or do inherent abnormalities predispose an infant to develop severe lower respiratory tract infection, in which case RSV is associated with the development of pulmonary sequelae (fig. 4⇓)? However, the majority of studies indicate that the infant, who develops severe RSV and subsequent wheezing, does have differences which predate the RSV infection. This supports the “association” hypothesis.

The strongest study design to prove causation would be a controlled clinical trial, in which subjects are randomly assigned to an intervention. If results for the intervention group and those for the control group differ, causation is demonstrated. A conclusive intervention study would need to be a major investigation, and the subjects would have to be monitored for many years. Such interventions include methods of preventing RSV disease by passive or active immunization. Until such large intervention trials can be carried out, it remains uncertain whether impaired lung function in children with a history of RSV bronchiolitis represents differences that predate the early RSV infection, or are caused by it 90.

Monoclonal antibodies

Administration of a humanized monoclonal neutralizing antibody to RSV, palivizumab, has proved to be an effective strategy to prevent RSV infection in premature and high-risk infants. Palivizumab has a strong binding affinity to the F protein and prevents the RSV spreading into the lower airways. The infant must receive the injections at regular monthly intervals throughout the RSV season to maintain optimal RSV protection. In the IMpact study 91, the incidence of RSV-related hospital admission was reduced by 55% in the studied risk groups (prematurity, <36 gestational weeks, or bronchopulmonary dysplasia). In absolute figures, the rate of hospital admission in the placebo group was 10.6% versus 4.8% in the palivizumab-treated group. The treatment is expensive, which limits large-scale use.

Antiviral treatment

Aerosolized ribavirin, a synthetic purine nucleotide derivative of guanosine, is the only antiviral drug available for treatment of severe RSV infections in infants and young children. It is virustatic, and its therapeutic efficacy in the acute infection is limited. Follow-up studies of the effect of ribavirin treatment on respiratory sequelae of RSV bronchiolitis show conflicting results. In a randomized placebo-controlled trial, Long et al. 92 were unable to demonstrate any statistically significant difference in pulmonary function 10 yrs after RSV bronchiolitis was treated with ribavirin or placebo. Similarly, Krilov et al. 93 found no statistically significant difference in reactive airway disease between treatment groups, 5–6 yrs after RSV bronchiolitis. In contrast, a prospective 5–7-yr follow-up study by Rodriguez et al. 94 of a placebo-controlled randomized trial suggested that ribavirin provides a long-term benefit versus placebo. A follow-up by Edell et al. 95 also suggested that ribavirin treatment of RSV bronchiolitis could reduce the prevalence of subsequent reactive airway disease.

Vaccines

The development of a vaccine for RSV remains an important goal in view of the clinical importance of the pathogen. A protective live attenuated vaccine that is administered at, or shortly after, birth would be ideal. Maternal immunization in the third trimester is also a possible alternative 96. Since young children have repeated RSV infections, the vaccine must produce better protection than is induced by natural RSV infection. Otherwise recurrent infections will not be hindered. Inhibiting antibodies from the mother are an obstacle to immunization of the baby, if the vaccine is given shortly after birth. Vaccines for RSV are under development 97. However, attempts to develop attenuated vaccines have, so far, not resulted in a commercially-available vaccine.

Although the development of effective virus vaccines is one of the major successes of biomedical research, an early vaccine for RSV has provided an example of unexpected, serious safety problems. Some children, who were immunized with an inactivated vaccine against RSV, developed a more serious infection than the nonvaccinated children when they came into contact with the wild-type virus. It is possible that the chemical inactivation had led to distortion of the immune response so that excessive production of IgE against one of the surface proteins occurred.

The use of recombinant DNA technology to produce RSV vaccine is now being studied. Such attempts concentrate on alterations of the extracellular domain of the F protein. Antibodies to the F protein are generally cross-reactive to both of the major RSV strains, A and B. Therefore, the F protein has been especially in focus in the development of recent candidate vaccines 98–101.

Can corticosteroid treatment of respiratory syncytial virus bronchiolitis diminish the risk of subsequent wheezing?

Several placebo-controlled studies have addressed the question whether corticosteroid treatment can influence the degree of respiratory sequelae after RSV bronchiolitis. Table 3⇓ summarizes these studies. The results are not consistent, although most follow-up studies report negative results 20, 21, 24, 102, 103. In a British study, early treatment with nebulized budesonide for 6 weeks neither decreased acute bronchiolitis symptoms (83% RSV positive) nor prevented postbronchiolitic wheezing during the following 6 months 20. In a Danish study, oral prednisolone treatment of children <24 months of age, hospitalized because of acute RSV infection, had no effect on outcome measures, either in the acute phase or in follow-ups, 1 month and 1 yr after admission to hospital 21. Similarly, a recent Dutch study found that oral prednisolone administered during the acute phase of RSV bronchiolitis did not prevent postbronchiolitis wheezing or asthma at the mean age of 5 yrs 102. In a British study, Fox et al. 103 studied the effect of administering inhaled budesonide for 8 weeks after hospital admission with acute viral bronchiolitis. Incidence of coughing or wheezing was not reduced for ≤12 months after bronchiolitis. Cade et al. 24 in another British study, also failed to find any short- or long-term clinical benefit of administration of nebulized corticosteroids in the acute phase of RSV bronchiolitis.

A few studies report positive long-term effects on postbronchiolitic wheezing after corticosteroid treatment. Kajosaari et al. 103 recently reported results indicating that inhaled corticosteroid treatment during and after the acute phase of RSV bronchiolitis in infancy (mean age 2.6 months) may have a beneficial effect on subsequent bronchial wheezing. The 117 children were followed up to 2 yrs after the bronchiolitis episode. In the group of children who had received inhaled budesonide for a week during the acute episode, 18% developed asthma; the figure for those treated with nebulized budesonide for 2 months was 12%. The corresponding figure for asthma development in the group that only received symptomatic treatment was significantly higher at 37%. The authors report that the children who seemed to benefit most from the treatment were those with atopy. In addition, a recent Swedish non-RCT follow-up study presents results indicating that inhalation of corticosteroids for 6–8 weeks may reduce subsequent asthma and severe respiratory morbidity in infants hospitalized for RSV infection (median age at hospitalization 2–3 months) 108. In conclusion, however, the majority of prospective placebo-controlled studies do not show any long-term beneficial effects of steroid treatment for RSV bronchiolitis.

The effect of anti-inflammatory treatment for older infants admitted to hospital with wheezing has also been investigated. Finnish studies have found that 4 months of treatment with nebulized budesonide reduced the recurrence of wheezing initially 89. However, the beneficial effect disappeared shortly after termination of the treatment (the children had a mean age of 10.6 months at inclusion and RSV had been detected in 26%) 104. In addition, the anti-inflammatory therapy had no influence on the occurrence of asthma 3 yrs later 105.