In the United States, pulmonary influenza infection occurs annually in 5-20% of the population with mortality in the range of 30,000 deaths. The recent 2009 influenza H1N1 pandemic illustrated the potential for higher infection rates, which were reported to be as high as 45% in certain age groups. The 2017-18 influenza season had the highest pediatric mortalities since the 2009 pandemic. Influenza infection is known to result in a broad spectrum of disease phenotypes in humans, although severe pneumonia is relatively rare. Despite this, severe disease often requires advanced supportive care in the young, including previously healthy children. Host factors involved in determining the outcome of influenza infection are unclear and children are known to be at higher risk of severe disease. First life exposure to influenza is also thought to dictate life-long immunity. Little is known about the effects of young age and gender on influenza responses and severity. This underscores the importance of understanding influenza pathogenesis in a pediatric population. Influenza pathogenesis is likely mediated in large part by exuberant inflammatory host responses in the lung. It is likely that predictive soluble inflammatory mediators are present in severe infection. Further, predictive biomarkers or mathematical models of influenza pneumonia severity would enhance clinical decision making and patient care. We propose that machine learning and mathematical modeling of host immune endpoints will define a molecular fingerprint of severe influenza pneumonia in juveniles. This hypothesis will be tested in two Aims.
Aim 1 will focus on characteristic molecular pathways related to influenza severity in juvenile animals, using outbred mice. We will utilize machine learning and new mathematical approaches for pathway and biomarker selection.
Aim 2 will test mathematical models of influenza pathogenesis to elucidate new mechanisms that drive lung injury. The overall goal of the proposed study is to identify novel biomarkers and mechanistic models of influenza pneumonia severity that can be applied to children. To accomplish this we will use a broad, exploratory, and unbiased approach. Candidate biomarkers and pathways would then be evaluated in future mechanistic and translational studies in mice and humans.
Pneumonia is the leading cause of death in children worldwide. There is a critical clinical need for biomarkers and models that predict influenza pneumonia severity in children. The goal of this study is to address this unmet need using an outbred mouse model of pediatric influenza pneumonia and well established mathematical approaches to inform future translational studies in children.
