Respiratory syncytial virus (RSV) is the leading cause of severe lower respiratory tract infections in infants, accounting for an estimated 3.6 million hospitalizations worldwide and more than 100,000 deaths in children under five each year. Although preventive antibodies have improved protection for vulnerable infants, treatment options once babies become seriously ill remain largely limited to supportive care.
Now, researchers at University College London (UCL) and Great Ormond Street Hospital have developed a human infant airway model that sheds new light on why RSV causes more severe disease in babies than adults. Published in Nature Communications, the study suggests that the infant airway itself actively amplifies damaging inflammation, highlighting the need for therapies that target both the virus and the immune response.
Severe RSV in infants has often been framed mainly as a story about an immature or overwhelmed immune system. Our data suggest that the infant airway tissue itself is actively shaping how aggressively immune cells respond.
—Claire Smith, University College London
"Severe RSV in infants has often been framed mainly as a story about an immature or overwhelmed immune system," Claire Smith, microbiologist at UCL and senior author of the study, told DDN. "Our data suggest that the infant airway tissue itself — the epithelium and the blood vessel layer beneath it — is actively shaping how aggressively immune cells respond."
Moving beyond animal models
Despite decades of research, translating promising RSV therapies from preclinical studies into successful treatments has proven difficult. Smith believes one reason is that traditional animal models struggle to capture the unique biology of the infant airway.
"Animal models have taught us a lot about RSV, but none of them fully replicate human infant disease," she said. "No single animal model captures the full picture: The immature but highly reactive immune system, the small and easily obstructed infant airway, and the specific way human epithelial and immune cells interact during infection."
To address this gap, the team built paired infant and adult airway models using primary human airway epithelial cells cultured at an air-liquid interface. They incorporated vascular endothelial cells beneath the airway tissue to recreate the blood-airway barrier and introduced primary neutrophils — the first immune cells recruited during infection — to observe how they migrated into infected tissue.
Creating the model required balancing the differing culture requirements of multiple primary human cell types while ensuring the short-lived neutrophils did not become activated during isolation and handling. "Each cell type needs its own specific culture conditions, and we had to test several different growth media and coatings before finding conditions the endothelial cells would tolerate without compromising the epithelial layer above them,” Smith explained.
Additionally, the team had to navigate the ethical issues of pediatric samples. “Unlike adult samples, infant airway cells and blood have to come from very young patients, often critically unwell children in intensive care, which places real ethical and logistical constraints on sample numbers.”
The infant airway drives inflammation
Using blood samples from infants hospitalized with RSV alongside their laboratory model, the researchers found evidence that neutrophils become highly activated during severe infection. Markers of neutrophil degranulation, including myeloperoxidase (MPO) and neutrophil elastase, were significantly elevated in infected infants compared with healthy controls. High levels of these enzymes are associated with tissue damage, inflammation, and oxidative stress.
When the team recreated this in their infant and adult airway models, the pediatric tissue recruited substantially more neutrophils than the adult tissue under identical experimental conditions. Those neutrophils also expressed higher levels of MPO despite no differences in barrier integrity between the two models.
"It shifts the focus," Smith said. "That means treating severe RSV effectively may require more than an antiviral to clear the virus; it may also require targeting the tissue-level signals that recruit and activate neutrophils, or the neutrophil degranulation pathway itself.”
More than viral load
The team compared two antiviral compounds, remdesivir and RSV604. Both drugs reduced viral replication to a similar extent. However, only RSV604 also reduced neutrophil MPO release.
"The finding that stood out most was from our drug-screening experiment," Smith said. "I hadn't anticipated that two drugs with similar mechanisms of action could affect neutrophil outcomes so differently, and it shows there is still a lot to learn about how these drugs act on human cells during infection."
The results highlight the limits of current antiviral screening strategies. "Current preclinical antiviral screening for RSV typically relies on viral load or plaque reduction as the main readout," Smith said. "Because our model incorporates the infant airway, the vascular barrier, and primary immune cells together, it lets us capture a therapeutic effect that a viral-load-only assay would miss entirely."
She added that future drug development should consider inflammatory biomarkers alongside antiviral activity. "Our take-home message from this work is that antiviral drug discovery should include neutrophil MPO reduction as a functional readout of efficacy, not just viral suppression."
A platform for future respiratory research
Although developed for RSV, the researchers believe the platform could become a versatile tool for respiratory disease research. The model could be adapted to investigate how pathogens such as influenza, rhinovirus, and bacteria interact with the infant airway while providing a more physiologically relevant system for therapeutic screening. By assessing a drug's effects on both viral replication and immune-driven inflammation, researchers hope it will help prioritize treatments with the greatest potential to improve outcomes in infants.
