The Respiratory Mucosa: Front and Center in Respiratory Syncytial Virus Disease

Infantile bronchiolitis is a major scourge of early childhood, and winter outbreaks fill the pediatric wards with wearisome regularity. Most cases are caused by respiratory syncytial virus (RSV), which was first isolated in 1956. Despite a vast amount of research in both human and animal models, a deep understanding of the inefficiency of protective immunity and, indeed, of the pathogenesis of RSV disease has been frustratingly slow to come by.

Most infants will be infected by RSV before their second birthday, with the risk of severe disease peaking at just 2 months of age. Despite the relative antigenic stability of the virus, reinfections with RSV occur throughout life. Studying disease in infants with primary disease presents considerable technical and logistical challenges; therefore, animal models (especially cotton rats, mice, and cows) have been widely used to enhance our understanding of primary infection and vaccine-enhanced disease. These models have been central in our efforts to understand the host immune response to RSV and the role of these responses in causing inflammatory bronchiolitis, but they do not recapitulate human disease in every detail.

Although animal models have advanced our understanding of the pathogenesis of bronchiolitis, a role for the type 2 immune responses (typical of atopic asthma and antihelminth immunity rather than antiviral responses) has been proposed (1). Type 2–biased responses have been suggested as a link between bronchiolitis and postbronchiolitic recurrent wheeze (2, 3), but it is not perfectly clear that prevention of RSV disease would modify the risk. Blanken and colleagues showed that prevention might substantially improve subsequent lung health in later childhood (4), but such an effect was not subsequently seen in large prevention studies in other settings (5).

With the prospect of even more effective and useful RSV immunoprophylaxis, novel therapeutics, and vaccines, it is increasingly apparent that observations from experimental models must be validated in human studies. In particular, the focus needs to be on the site of infection: the ciliated mucosal respiratory epithelium. To this end, in a study presented in this issue of the Journal, Vu and colleagues (pp. 1414–1423) (6) sought parallels to the previous finding that the airways of neonatal mice had increased numbers of type 2 innate lymphoid cells (ILC2s) after RSV infection (7), as was also observed in neonatal mice with rhinovirus infection (8). They compared children admitted to pediatric intensive care wards (who mostly required mechanical ventilation) with those admitted to general wards.

When they examined nasopharyngeal cells (9) collected from the subjects, they found that the frequency of ILC2s was indeed increased in infants with more severe disease. In addition, they found that levels of the ILC2-activating alarmin IL-33 and the ILC2-secreted type 2 cytokines IL-4 and IL-13 were also elevated in infants with severe disease. In agreement with other recent studies (10), levels of IFN-γ were paradoxically lower in patients with severe disease, supporting an association between enhanced severity and diminished antiviral immunity.

Vu and colleagues also provide further support for the association between premature birth and RSV severity. This association is well known, but the mechanisms underpinning it seem to be complex. Lower maternal antibody titers at birth predispose to infection, and smaller airway size may increase the likelihood of airway plugging, a feature of severe disease (11). However, another likely contributor is the rapid development of the respiratory immune system in early life, when ILC2s are particularly abundant (12). Indeed, the present analysis by Vu and colleagues demonstrates that among infants hospitalized with RSV, ILC2s were most abundant in those less than 3 months of age. This may help explain the high rate of severe disease at 2 months of age, when ILC2s may be particularly prevalent in the airway. Vu and colleagues also found that IL-4 levels were higher in infants less than 3 months of age, in agreement with previous studies (13).

Given the possibility that ILC2s and T-helper cell type 2 (Th2) mediators might play a pathogenic role, further studies of the dynamics of ILC2 abundance in the airway are needed. Many issues remain to be addressed, such as the effects of ILC2s in the airway before infection. These cells are elevated in normal neonatal mice (7, 12), but is this the case in human infants? If so, this might provide an added thread to our understanding of the association between age and severity of disease. To explore this possibility, Vu and colleagues incorporated gestational age into a multivariable logistic regression model and demonstrated that both younger gestational age and higher airway IL-4 levels were associated with severity. However, it is possible that these factors reflect an existing immune profile of the airway rather than an acute response to infection.

Although ILC2-mediated enhanced type 2 immunity to RSV infection may reciprocally diminish antiviral type 1 immunity (including IFN-γ levels, and potentially CD8⁺ tissue resident memory cell activity in reinfection [14]), a consequential increase in viral load is not seen in severe disease (10, 15). Indeed, IFN-γ and IL-33 have reciprocal effects on respectively inhibiting and enhancing ILC2 activity (16). In severe RSV disease, where IFN-γ levels are relatively low and IL-33 levels are relatively high, activation of lung-resident ILC2s (and/or migration of ILC2s to the lung) apparently results. It remains unknown whether type 2 inflammation contributes to the control of RSV load in infants (regardless of severity) or whether it is simply pathologic. By improving our understanding of ILC abundance and activity in the developing healthy neonatal respiratory tract, we may be able to gain additional insights into the contribution of these cells to the host response to RSV infection.

Future studies of the mucosal immune response throughout infection and into convalescence are warranted to answer these questions. Our incomplete understanding of the relative contributions of viral replication and host inflammatory responses continues to hamper the development of novel therapies, possible immune modifiers, and vaccines. An obvious question remains unanswered: if Th2 responses are important, why is RSV bronchiolitis characterized by airway neutrophilia (17) rather than eosinophilia?

Vu and colleagues are to be applauded for tackling these difficult questions head on. By focusing on the hard-to-measure responses in the airway mucosa, they were able to gain novel insights that enable a nuanced interpretation of mechanistic animal studies, and demonstrate the potential importance of ILC2s and type 2 cytokines in RSV disease and its sequelae. These new findings support the concept that type 2–driven immunopathology plays a part in severe RSV disease, but much remains to be discovered about the dynamic, ILC-rich airway in early life and during mucosal infections.