Definitions

Bronchiolitis is an acute communicable disease predominantly manifesting in infancy and characterized by cough, coryza, fever, grunting, tachypnea, retractions, inspiratory crackles, expiratory wheezing, and air trapping. The first (and the most severe) episodes occur most frequently in infants aged 2 to 6 months. Infectious asthma is a term that generally refers to infection-induced wheezing occurring beyond infancy. Nonetheless, a patient may experience bronchiolitis during the first months of life and a recurrent episode caused by the same virus in the second year, suggesting an identical underlying nature of the two illnesses. Certain viruses (e.g., respiratory syncytial virus [RSV] and influenza viruses) seem capable of causing bronchiolitis in all children; in contrast, other agents (rhinovirus in particular) seem to induce wheezing primarily in atopic children.

History

Although all practicing pediatricians today are familiar with the term bronchiolitis and associate it with an acute clinical illness with the signs and symptoms of airway obstruction, it has been recognized in the medical literature for only a relatively brief period. Acute bronchiolitis finally was listed as a heading in the sixth edition of the Textbook of Pediatrics, published in 1954; even at that time, bronchiolitis was associated with interstitial pneumonitis.

In contrast to the delayed textbook recognition of bronchiolitis, good clinical descriptions were presented in journals. In 1941, Hubble and Osborn published “Acute Bronchiolitis in Children,” in which they described an epidemic of bronchiolitis involving 50 hospitalized children. Pratt, in 1944, and Nelson and Smith, in 1945, published excellent clinical articles. Studies in the late 1950s and early 1960s established the etiologic association of RSV and other viruses with acute bronchiolitis.

Etiologic Agents

RSV is the major cause of bronchiolitis in infancy and virtually the only important etiologic consideration when major epidemics are occurring. Other agents cause smaller annual outbreaks. The frequency of infectious agents in the overall cause of bronchiolitis is shown in Table 21.1 . These data were compiled from 12 reports in which respiratory illness in children was observed over an extended period, and in most instances numerous clinical illness categories were being studied. During major epidemics in the colder months in temperate climates, virologic and serologic studies indicate RSV as the cause in 80% or more of the cases, especially in the more severe ones. In nonepidemic situations, more than 50% of the isolates or instances of serologic evidence of infection are infectious agents other than RSV. Human metapneumovirus becomes a prominent cause of bronchiolitis as the RSV season ends. In nonepidemic situations, more than 50% of the isolates or instances of serologic evidence of infection are infectious agents other than RSV, including parainfluenza viruses, and rhinovirus.

Children surviving infantile bronchiolitis often have recurrent episodes of wheezing that, on clinical grounds, seem to be precipitated by upper respiratory tract infections. Studies designed to determine the infectious agents responsible for these repeated wheezing episodes have yielded variable results, presumably because they involve patients of different ages observed during different seasons of the year.

A comprehensive study based on more than a decade of observation of children from birth through 15 years of age revealed that viruses and Mycoplasma pneumoniae are the most common etiologic agents ( Table 21.2 ). RSV is the most common cause of recurrent, wheezing-associated respiratory illness occurring in young children, including children who previously experienced bronchiolitis, which strongly suggests the identical nature of infection-related wheezing episodes occurring in children of different ages. With increasing patient age, rhinoviruses and M. pneumoniae account for most infection-induced wheezing episodes. The parainfluenza viruses, metapneumoviruses, and influenza viruses are important causes of such wheezing episodes throughout childhood, whereas coronaviruses and adenoviruses are less common causes in infancy. Human bocaviruses are more recently identified causes of bronchiolitis, and their relative importance as etiologic agents is being determined. RSV remains an important cause of wheezing in adolescents and can cause airway obstruction in middle-aged and elderly individuals.

The role of bacteria as primary agents, synergistically active participants, or secondary invaders in bronchiolitis has interested investigators for many years. No evidence has been found of a primary role for bacteria alone in bronchiolitis of infants. The data on synergistic viral-bacterial infections also are unconvincing. Wood and colleagues in 1954 and Sell in 1960 reported results of studies in which bacteriologic and serologic evidence tended to associate Haemophilus influenzae infections with bronchiolitis. Some of the children from these studies later were shown serologically to have had RSV infection. Later studies showed evidence of mixed viral-bacterial infections in bronchiolitis and other lower respiratory tract illnesses.

These serologically defined mixed infections were found no more commonly among patients with bronchiolitis than among patients with mild upper respiratory tract disease, whereas RSV was recovered 10 times as often from patients with lower respiratory tract disease as from patients with upper respiratory tract disease alone. Although debate continues about whether the isolation of H. influenzae from the airway of infants with bronchiolitis represents secondary infection, most data indicate that a significant role for mixed viral-bacterial infection in bronchiolitis of infancy is unlikely to exist. Bacterial superinfection is an uncommon occurrence in bronchiolitis (at least in developed countries). Generally after a diagnosis of RSV or parainfluenza virus infection has been established, antibiotic courses started empirically can be safely stopped. Nevertheless while bacterial superinfection is an uncommon occurrence in bronchiolitis in developed countries, there is evidence that bacterial superinfection of RSV bronchiolitis occurs in developing countries. Initiation of pneumococcal vaccination programs has led to reduced rates of hospitalization for RSV (and influenza) infections, suggesting that superinfection with pneumococci led to more severe forms of RSV infection.

The association of asthma and infectious illnesses was appreciated throughout the 20th century. Initial reports tended to relate bacterial agents to wheezing; in 1909, Carmalt-Jones reported improvement of patients with asthma in association with injections of a bacterial vaccine. Throughout the first half of the 20th century, “bacterial allergy” was the main consideration in infection-related wheezing, and controversy raged about the therapeutic merits of bacterial vaccines. Despite widespread use of bacterial vaccines in the treatment of infectious asthma, no evidence exists that bacterial organisms of the normal flora precipitate asthmatic attacks. More recent, large retrospective reviews have shown the rarity with which bacterial infection complicates bronchiolitis, even in severe cases. Generally after a diagnosis of RSV, metapneumovirus, rhinovirus, or parainfluenza virus infection has been established, antibiotic courses started empirically can be safely stopped.

Epidemiology

Bronchiolitis predominantly is a disease of infancy, although an identical disease occurring in older children arbitrarily may be referred to as another diagnostic entity. In a study involving 1148 children, the peak age of incidence was 2 to 6 months, with more than 80% of the cases occurring during the first year of life. In a small study of families in or near Houston, Texas, rates of RSV infection were 68.8 per 100 children in the first year of life and 82.6 per 100 children in the second year, including many reinfections; larger studies have not been completed. Lower respiratory tract illness was caused by RSV in 22.4 of 100 children in the first year of life. Of all RSV infections occurring in children younger than 12 months of age, one-third were accompanied by lower respiratory tract illness. Although the attack rate of RSV decreased with age, the frequency with which lower respiratory tract disease occurred among individuals infected remained constant ( Table 21.3 ), at least until they reached 4 years of age. In Tucson, Arizona, a study of 1179 children enrolled in a health maintenance organization found that the rate of lower respiratory tract illness in the first year of life was 32.9 episodes per 100 children; 60% of these episodes were diagnosed as bronchiolitis.

A study of the Tennessee Medicaid population found that the frequency of hospitalization for bronchiolitis was 4.4 and 1.5 per 100 child-years of observation for infants infected in the first and second 6 months of life, respectively. Approximately 123,000 infants younger than 1 year of age are hospitalized annually with bronchiolitis, with RSV infection accounting for 51,000 to 82,000 of these hospitalizations. Including cases diagnosed as pneumonia, RSV infection results in 70,000 to 120,000 hospitalizations.

The yearly incidence of hospitalization for bronchiolitis increased from 12.9 per 1000 infants in 1980 to 31.2 per 1000 in 1996, suggesting an increased severity of illness developing during this interval. More frequent hospitalization may also be a result of an increased use of pulse oximetry to determine the need for hospitalization. In concert with this, a subsequent study showed that the rate of RSV hospitalizations was stable from 1997 to 2006, at a rate of 132,000 to 172,000 hospitalizations annually in children younger than 5 years old in the United States. More recently, however, the rate of hospitalization for RSV infection has plateaued, but the severity of illness among hospitalized cases seems to have increased. This suggests that hospitalization is being progressively avoided among infants with milder forms of illness.

Epidemic bronchiolitis caused by RSV is markedly seasonal in temperate climates, with peak activity occurring from January to April (Northern Hemisphere) annually and with virtually no activity seen from August to October. Sporadic bronchiolitis cases caused by other agents are seen throughout the year. Regional differences can be striking. In Miami and Puerto Rico, RSV activity occurs year round, with a peak of activity from July to October. Epidemics usually begin in October in other southern states, although still earlier than in the Northeast. Farther north (Winnipeg, Canada, and Alaska), the duration of the RSV season again becomes longer, and cases occur essentially year round. Near the equator, temperature and humidity are positively associated with the number of bronchiolitis cases. In contrast, farther from the equator (in either direction), temperature and humidity are inversely associated with bronchiolitis cases. Ultraviolet radiation is negatively associated with bronchiolitis in several regions. In contrast to yearly epidemics of RSV infection in the United States and United Kingdom, with identical peak months of occurrence each year, RSV occurs in biennial epidemic cycles in Croatia, Switzerland, Sweden, Finland, and Germany, with peaks alternating between December and January in 1 year and March and April in alternate years. No explanation has been established for the biennial occurrence of these epidemics.

Bronchiolitis occurs more frequently in boys; the male-to-female ratio is approximately 1.5 : 1. Crowding may be a major determinant of hospitalization rates for lower respiratory tract illness caused by RSV, with the incidence of hospitalization for infants 1 to 3 months old residing in rural, urban, and heavily industrialized areas of Great Britain being 1 in 80, 1 in 60, and 1 in 40, respectively. The figures for all children younger than 5 years of age were 1 in 714, 1 in 588, and 1 in 227, respectively. Studies from Tucson, Arizona, indicate that many socioeconomic factors, including absence of breastfeeding, low level of maternal education, and exposure to cigarette smoke, are associated with an increased risk for developing lower respiratory tract infection at the time of infection with RSV. The greatest risk was conferred by sharing sleeping quarters with two or more individuals.

Clinical Presentation

Acute bronchiolitis occurs most commonly in infants 1 to 12 months of age. In most instances, the patient’s history reveals exposure to an adult or older child with a common cold or other trivial respiratory tract infection. Occasionally, as in the daycare setting, the child is exposed to other children with more marked respiratory illness. After exposure, the incubation period is 5 to 7 days. The initial signs include copious nasal discharge (often serous in the early stage), cough, irritability, poor feeding, and vomiting in some cases. Slightly more than 50% of infants have fever, with rectal temperatures ranging from normal to 40.6°C (105.4°F) (mean, 39°C [102°F]). Nasal congestion with tenacious secretions and progressive cough and dyspnea dominate the clinical picture.

Symptoms of upper respiratory tract infection persist for a few days, and the onset of lower respiratory tract infection usually is precipitous, with the time of the onset of wheezing often being recognizable from the caretaker’s description of the illness. The maximum severity of illness generally is attained within 24 to 36 hours of the first signs of lower respiratory tract illness, and fewer than 2% of hospitalized infants deteriorate clinically enough to require transfer to intensive care units. Apnea may occur and may be sufficiently severe to require mechanical ventilation.

At the time of hospital admission, all patients have cough and evidence of respiratory distress. The pulse is rapid, and the respiratory rate usually is 40 to 80 breaths per minute. The breathing is labored, with flaring of the alae nasae; grunting; abdominal breathing; and supraclavicular, subcostal, and intercostal retractions. The degree of retraction of the lower chest wall may be an accurate indicator of the severity of the illness. Wheezing is best heard by auscultation over the upper anterior chest but often is audible without the use of a stethoscope (particularly prominent over the upper, anterior chest), and the chest is often hyperinflated. Hyperresonance may be detected on percussion, and auscultation generally reveals harsh rhonchi, high-pitched or low-pitched expiratory wheezes, or fine inspiratory crackles. Occasionally wheezing is not audible despite other evidence of airway obstruction, due to low airflows. A prolonged expiratory phase of breathing may occur, suggesting the presence of a more severe degree of illness. Cyanosis also occurs in severe cases.

Other findings include a mild conjunctivitis in one-third of cases, pharyngitis of varied severity in approximately half of affected infants, and otitis media in 5% to 15% of cases. The abdomen frequently appears distended, and the liver and spleen may be palpable; the organs are not enlarged but are pushed down because of hyperexpansion of the lung.

The duration of the hospital course of bronchiolitis varies. Significant improvement occurred in half of the cases within 2 days in one large study. In the same study, approximately one-third of the cases had a gradual course without evidence of clear-cut improvement at any one time; 71% of the patients were afebrile by the third hospital day. In another study done more than 25 years later, the average duration of hospitalization was 3.4 days. Longer stays were required for infants with initial oxygen saturations of less than 90% and infants younger than 6 weeks at the onset of illness. Reasonable criteria for admission in otherwise healthy infants seem to include hypoxia (i.e., oxygen saturation of <90% to 92%), age younger than 6 weeks, and a degree of respiratory distress sufficient to reduce fluid intake to inadequate levels. Other criteria include apnea, immunodeficiency, premature birth, and the presence of significant underlying heart or lung disease. Most patients can be discharged from the hospital within 2 to 3 days after admission, although mild wheezing still may be present for several days or even weeks thereafter.

In 121 hospitalized patients, the total white blood cell count was less than 12,500/mm ³ in 74% and in 15 determinations had more than 60% neutrophils. In another study, a mean leukocyte count of 16,000/mm ³ was determined for children with lower respiratory tract disease, including bronchiolitis, as was an increased percentage of band form neutrophils compared with that of a control group. As in most infections, eosinophil counts in peripheral blood are reduced at the time of acute RSV infection. Nonetheless, some patients maintain detectable eosinophilia in peripheral blood; these infants may be more likely to develop childhood asthma.

Although impaired oxygenation is usually present in infants with respiratory distress, cyanosis may not be evident, even in the presence of markedly reduced oxygenation. The respiratory rate is related inversely to the degree of oxygenation except when respiratory failure is imminent. In mild to moderate cases, carbon dioxide retention does not occur because the alveoli that are functioning can compensate for alveoli that are not ventilated. In severe disease, the blood pH is low and the Pa co ₂ is elevated. The technique of pulse oximetry has obviated the need for arterial blood gas sampling except perhaps in severe cases in which hypercarbia is a concern.

In certain patients, clinical findings (e.g., degree of chest wall retractions, wheezing) often are out of proportion to the degree of hypoxia as measured by pulse oximetry. Infants with marked dyspnea must be evaluated carefully because respiratory failure may occur precipitously, despite their reassuring oximetry readings. Once infants are hospitalized and stabilized, subsequent deterioration to the point of requiring intensive care is unusual, occurring in less than 2% of cases. Respiratory rates of 63 breaths per minute and oxygen saturations less than or equal to 88% on admission were weakly associated with such eventual deterioration in one study.

The radiographic appearance of the chest in bronchiolitis varies considerably. Anteroposterior radiographs may be normal in mild cases or may reveal hyperinflation in the absence of hypoxia. In moderate to severe illness, radiographs often appear exceptionally clear because of hyperinflation. The diaphragms often are flattened or depressed. The costophrenic angle is less acute, and the hilar vascular shadows are stretched. Frequently areas of atelectasis may give the appearance of pneumonitis, although true consolidation is a rare occurrence. The heart usually appears small. On the lateral radiograph, the diaphragm is depressed markedly, and reversal of the normal convexity frequently may be seen. The anteroposterior diameter of the chest is increased.

With recurrent wheezing episodes, the prodrome may be considerably shorter in duration with little or no fever and minimal coryza. In these children, differentiation of infection-induced wheezing from more conventional asthma becomes essentially impossible clinically.

Pathophysiology

The pathophysiology of bronchiolitis deservedly has been the focus of numerous investigations. Several original theories can be discounted, whereas others warrant further study. Pathologic examinations of the lung in bronchiolitis or human airways in organ culture reveal necrosis of the respiratory epithelium with destruction of the ciliated layer; mononuclear cell invasion of the peribronchial tissues and alveolar interstitium; edema of the submucosa and adventitia; and obstruction of small airways with dense plugs consisting of dying epithelial cells, fibrin, and inflammatory cells. In contrast to asthma, mucin (periodic acid–Schiff–positive material) usually is absent in fatal bronchiolitis. An associated interstitial pneumonitis often is present, and patchy areas of atelectasis frequently are noted.

The predominant cell types present in the lung of fatal cases are macrophages and neutrophils. Lymphocytes bearing CD4 (helper cells), CD8 (cytotoxic cells), and CD56 (natural killer cells) antigens are almost absent in fatal bronchiolitis cases. In contrast to the findings in fatal cases, CD8 lymphocytes were present in the lung tissue obtained from a nonhospitalized child with a mild lower respiratory tract infection caused by RSV who died in an automobile accident. Even in surviving infants, CD8 lymphocyte counts are less than 1% of the total cells in bronchoalveolar lavage fluids and only appear in respiratory tract secretions of subjects with severe forms of disease after illness has largely resolved. These findings suggest that CD8 lymphocyte responses are necessary for early control of RSV infection and that severe disease is perhaps related to an absence of CD8 responses. Cytokines characteristically released by CD4 and CD8 cells (interleukin [IL]-2, IL-4, IL-5, IL-13, and interferon [IFN]-γ) are present only in low concentrations in nasopharyngeal and tracheal secretions of infants with severe lower respiratory tract infections caused by RSV. Microarray analysis of gene expression during acute RSV infection confirms these findings. Genes related to innate IFN function and neutrophil function were among the most overexpressed genes in infants infected with RSV. Genes regulating T and B cells (components of the adaptive immune response) were among the most underexpressed genes.

Staining for markers of apoptosis (caspase 3 and Fas) is positive in infected epithelial cells, suggesting that recovery from primary infection depends on the antiviral activity of phagocytic cells, induction of apoptosis, and perhaps the activity of other innate immune mechanisms. Strong innate immune responses occurring early in the course of infection may limit the eventual degree of illness. Rhinovirus bronchiolitis appears to occur more frequently in subjects with preexisting atopic dermatitis and eosinophilia, suggesting that allergic pathways may contribute to bronchiolitis caused by this virus.

Recovery from bronchiolitis apparently is complete histologically, although plugging of the airway still was prominent in an infant 5 weeks after having an acute episode of bronchiolitis. This infant had experienced an apparent full clinical recovery before eventually dying of acute pneumococcal pneumonia.

Infants may be particularly prone to the development of severe illness as a result of infection of the small airways for many reasons, including the small diameters of their airways. The infant lung is deficient in collateral alveolar ventilation through the pores of Kohn, which develop only in later life. Atelectatic areas therefore cannot be re-expanded readily. Other studies have shown that the small airways in children younger than age 5 years contribute five- to sevenfold more to total airway resistance than do small airways of adults. Viral infections involving small airways in young children are more likely to present as more serious clinical manifestations than similar infections in adults. Nonetheless, these abnormalities exist in all infants, whereas most infants infected with RSV do not develop lower respiratory tract illness. This suggests that host or environmental factors may be involved in determining the pathogenesis of RSV infection. Environmental factors are listed in Table 21.4 , and potentially important host factors are described later.

Another factor that may predispose to the development of lower respiratory tract illness after RSV infection is the infecting dose of virus. Some studies show that infants with greater quantities of RSV in their nasopharyngeal secretions are more likely to exhibit severe illness, although the results of other studies show no correlation with the amount of virus present in the respiratory tract with the severity of illness. In fatal cases, abundant virus is evident in epithelial cells plugging the airway lumen by immunohistochemical staining. The fact that crowding is associated with greater risk for developing lower respiratory tract infection also suggests the importance of the initial inoculum in causing disease.

Another risk factor for the development of lower respiratory tract infection may be the relative diameter (or degree of intrinsic constriction) of the airway. When pulmonary function testing is completed on healthy infants before lower respiratory tract infection occurs, certain infants have lower air flows in smaller airways (and presumably narrower airways) than do other infants. When observed prospectively, these infants are more likely to develop wheezing early in life than are infants with better airflow. These abnormalities of lung function no longer are associated with an increased risk for wheezing after the child reaches 3 years of age. Instead, evidence of atopy becomes the principal risk factor for recurrent wheezing.

Prospective studies of infants who had bronchiolitis in infancy demonstrate reduced airflow in small airways in later childhood. Studies of adolescents reveal reversibility of this restriction of airflow after receiving bronchodilator treatment, indicating the airways were constricted, but not stenotic, in early life. Several studies have shown that the airways of infants are intrinsically more reactive to bronchospastic stimuli than are airways of older children, particularly children from families with asthma. How long this increased reactivity persists is unknown, and no study has yet shown that infants with greater degrees of hereditary airway hyperreactivity are more likely to develop bronchiolitis or recurrent virus-induced wheezing. In later childhood, repeated occurrences of viral infections are necessary to sustain this increased reactivity.

In addition to the hereditary airway hyperreactivity, viruses themselves may induce increases in reactivity. Viral infections that clinically appear to be restricted to the upper respiratory tract in adults and children nonetheless result in transient, increased constrictive responses of the airway to a variety of stimuli, including histamine, irritants, and other agents, and small airway dysfunction. Whether these minor changes contribute to the airway obstruction observed in bronchiolitis is doubtful, in that they are of much lesser magnitude than is the markedly increased reactivity observed in asthmatic individuals after exposure to allergens. These virus-induced changes were observed in all infected individuals, including subjects who did not experience wheezing at the time of the virus infection. If airways already are hyperreactive in infants, on a hereditary basis or as a result of a preceding viral infection, the subsequent stimulus of RSV infection may be sufficient to cause airway obstruction.

Immunologic deficits may be expected to lead to more severe forms of virus-induced respiratory illness. Studies of the antibody response to RSV infection in serum and in respiratory secretions show, however, that the nature of responses generally is similar among patients with bronchiolitis or simple upper respiratory tract illness alone caused by this agent. Antibody-directed cellular cytotoxicity expressed against tissue culture cells infected with RSV also is similar in patients with all forms of illness caused by RSV. Although investigators have shown that cells infected with RSV can activate complement through classic and alternative pathways, studies suggest that in vivo activation of the complement cascade occurs with equal frequency among patients with all forms of illness caused by the virus. Repeated RSV infections in early life may occur because antibody formed after primary RSV exposure has a low ability to bind to neutralizing epitopes on the virus.

The concept that serum IgG antibody to RSV, acquired by vaccination or transplacentally, might sensitize the host has been dispelled by results of studies showing that severe bronchiolitis may occur in the absence of circulating antibody and that titers of maternal antibody correlate with protection against infection caused by the virus. Circulating immune complexes appear as frequently in mild RSV-related upper respiratory infections as in RSV bronchiolitis.

Lymphocyte hypersensitivity has been suggested to play a role in the development of severe bronchiolitis. Original field trials showed that a formalin-inactivated RSV (FI-RSV) vaccine induced humoral and cell-mediated immune responses to the virus. Nonetheless, vaccinated subjects manifested more severe forms of illness than did unvaccinated controls when subjects in each group subsequently were infected naturally. Cell-mediated immune responses to viral antigen were greater among recipients of this FI-RSV vaccine compared with control subjects who previously had experienced natural infection. The resulting disease in vaccine recipients was similar in form to bronchiolitis, and the idea that natural RSV bronchiolitis is a consequence of lymphocyte hypersensitivity to the virus has persisted.

Nonetheless, it seems unlikely that the illness that followed FI-RSV vaccination is mediated in the same way as that following natural bronchiolitis. The two FI-RSV vaccine recipients who died after having subsequent natural infection had numerous lymphocytes and eosinophils present in the lung, which is not the case after natural RSV infection. Infants surviving RSV infection after receiving the FI-RSV vaccine had high eosinophil counts in peripheral blood, which also is unusual in natural RSV infection. In addition, cytotoxic T-lymphocyte activity is not detected easily in infants after bronchiolitis, and more recent autopsy studies have shown the virtual absence of CD4 and CD8 antigen-positive cells in the lung. Although the development of cell-mediated cytotoxic antiviral responses may be protecting individuals with milder disease, little reason exists to contend that lymphocyte hyperresponsiveness contributes to severe bronchiolitis.

Immediate hypersensitivity to viral antigens has received much consideration as a potential contributing factor in bronchiolitis. The association of a family history of asthma with the development of bronchiolitis in infancy remains controversial. Production of virus-specific IgE and subsequent release of mediators of bronchoconstriction have been documented, however, in infants hospitalized with bronchiolitis caused by RSV and the parainfluenza viruses. In bronchiolitis caused by RSV, the quantities of virus-specific IgE produced and histamine present in respiratory secretions correlate with the severity of illness as measured by degree of Pa o ₂ . Leukotriene C ₄ is a product of mast cells and eosinophils primarily, although also epithelial cells and macrophages, and is a potent stimulant of airway smooth muscle constriction and mucus secretion. This mediator is released into the airway in acute bronchiolitis, as is the bronchoconstrictive agent histamine. In addition, prostaglandins and metabolites can be detected in secretions and blood after bronchiolitis. Thus several mediators of airway obstruction that are released in response to the interaction of IgE with antigens are released in subjects with bronchiolitis.

A possible role for eosinophils in the pathogenesis of bronchiolitis is supported by the finding of increased concentrations of eosinophil cationic protein in secretions of infants with bronchiolitis and (among high responders) an overall correlation of concentrations of this protein with the degree of hypoxia. Peripheral blood eosinophil counts are depressed during the acute phase of most infectious diseases, including RSV infection. Nonetheless, eosinophil counts in peripheral blood are higher in infants with bronchiolitis than in infants with upper respiratory tract illness only, particularly in boys. This finding is notable because boys generally have more severe forms of bronchiolitis. A contrasting viewpoint is that eosinophils contribute to the clearance of viruses after infection. This view is supported by the fact that eosinophils contain enzymes with ribonuclease activity, which inactivates RSV. IgE-dependent and eosinophil-dependent mechanisms may be important in recovery from viral infections, much as they are in recovery from parasitic infections.

The airways of asthmatic individuals are infiltrated by T lymphocytes (predominantly T-helper lymphocytes) and eosinophils. T-helper cells can be classified as type 1 (T _H 1), which produce primarily IFN-γ and IL-2, or type 2 (T _H 2), which produce IL-4, IL-5, IL-13, and others. This finding is important because IL-4, IL-5, and IL-13 are important factors in promoting IgE synthesis and eosinophil migration. Asthma is thought to be induced by a T _H 2 bias in the airway. An attractive hypothesis is that RSV infection induces airway obstruction and wheezing by inciting the same type of T _H 2 responses as observed in asthma. However, although IFN-γ and IL-4 are found in respiratory secretions of infants with RSV bronchiolitis, IL-5 and IL-13 rarely are detectable in serum or secretions, suggesting that bronchiolitis is not characterized by a T _H 2 lymphocyte bias. RSV infection in infancy does not induce a long-term Th2 lymphocyte cytokine bias lasting into later life.

Chemokines are proteins released by airway epithelial cells, inflammatory cells, fibroblasts, and other cell types that are chemotactic for leukocytes. A failure to initiate strong chemokine responses at the onset of RSV infection has been linked to more severe disease outcomes, possibly because innate immune mechanisms are not activated in a timely fashion. Chemokines represent an alternative to T _H 2 lymphocytes in terms of initiating the inflammatory response in asthma and bronchiolitis. However, although numerous studies of the chemokine response in RSV bronchiolitis have appeared, no chemokine has been consistently associated with worse outcomes. In contrast, the release of IFN-γ has been more consistently correlated with better outcomes.

Differential Diagnosis

A tendency exists to attribute all infantile respiratory distress occurring after the immediate neonatal period to bronchiolitis. The list of causes of infantile dyspnea consists of conditions that are associated with upper and lower airway obstruction, however. Recognition of upper airway obstructive disease should cause little difficulty because the problem is one of distress with inspiration rather than air trapping. Illnesses that cause lower airway obstructive disease and diseases that suggest this problem are listed in Box 21.1 .

The differential diagnosis of allergic disease causes the most difficulty. Generally the first episode of allergic respiratory disease, when associated with infection, cannot be separated from bronchiolitis by any objective measures. Anatomic defects such as vascular rings can cause obstruction of the airway at many locations; inspiratory or expiratory distress, or a combination, can occur. Frequently a child with an anatomic defect does not have any detectable difficulty until a trivial respiratory tract infection occurs, which complicates the diagnostic picture.

Foreign bodies should be considered even in very young infants. Gastroesophageal reflux disease has become recognized as a frequent cause of wheezing in young infants. Because of its obvious therapeutic implications, bacterial pneumonia is the most important differential consideration, although wheezing occurs rarely in association with bacterial pneumonia.

Diagnosis

Because bronchiolitis and infectious asthma are clinical diseases with arbitrary boundaries and multiple etiologic agents, outlining a method for establishing specific diagnosis is difficult. When illness is epidemic, RSV usually is the cause. In nonepidemic situations, a careful history and appropriate laboratory studies and radiographs should be considered to exclude the other differential diagnostic possibilities that are listed in Box 21.1 .

A specific etiologic diagnosis can be made by the isolation of virus from the nasopharynx. The diagnostic virologic facilities of many medical centers enable the isolation in tissue culture of RSV, parainfluenza viruses, influenza viruses, adenoviruses, rhinoviruses, enteroviruses, and herpesviruses. The use of the shell vial technique has been applied to RSV infection with success. Until recently, rapid detection of viral antigen of RSV directly in nasopharyngeal secretions by commercially available (e.g., enzyme-linked immunosorbent assay) or fluorescent antibody techniques has been the method of choice in most laboratories. The accuracy of these techniques often is superior to that of standard cell culture (at least in infants) because antigens remain stable under transport conditions that inactivate live virus. Gradually the technique of amplifying small quantities of RNA of respiratory viruses simultaneously from clinical specimens (polymerase chain reaction) is replacing earlier methods. This is because the sensitivity of these assays in detecting minute quantities of RSV is far greater than that of other methods, and simultaneous testing for 20 or more infectious agents is possible in a small clinical specimen.