Influenza and the Respiratory Microbiome

Abstract

Despite the availability of vaccines, influenza causes approximately 3-5 million cases of severe illness and 400,000 deaths each year. Prevention efforts might potentially be strengthened by harnessing the host microbiome, which plays an important role in maintaining human health by promoting host immunity and colonization resistance. Although vaccines are the best available means of prevention, vaccine effectiveness has been low to moderate in recent years and vaccine coverage remains low, especially in low- to middle-income countries. Exploring the relationship between influenza virus and the respiratory microbiome may contribute to alternative strategies of prevention. This dissertation explores the relationship between influenza virus and the respiratory microbiome. In chapter 2, we describe our current understanding of respiratory virus-bacteria interactions using systematic and targeted literature searches. We explore whether respiratory viruses can place selective pressures on bacteria in the upper respiratory tract. Further, as colonization in the upper respiratory tract is a necessary precursor for many respiratory pathogens, we explore whether virus-associated changes in the upper respiratory tract microbiome can influence the etiology of bacterial pneumonia. We found strong biological support for a link between respiratory viruses, the upper respiratory tract microbiome, and bacterial pneumonia. However, we found a lack of longitudinal studies among human populations that examined all three components. To address this knowledge gap, we used a household transmission study of influenza in Nicaragua to explore potential relationships between influenza and the respiratory microbiome. In chapter 3, we examine whether the respiratory microbiome mediates susceptibility to influenza virus infection and characterize structural changes to the respiratory microbiome during influenza virus infection. We used Dirichlet multinomial mixture models to assign nose/throat samples to bacterial community types and generalized linear mixed effects models which account for clustering by household. We found a single community type associated with decreased susceptibility to influenza. Further, we found high rates of change in the microbiome structure following influenza virus infection as well as among household contacts who were never infected with influenza during follow up. In chapter 4, we use secondary cases from the Nicaraguan household transmission study to investigate whether the respiratory microbiome impacts influenza symptomology and viral shedding. We used generalized linear mixed effects models to examine the presentation of symptoms and viral shedding. Further, we used accelerated failure time models with a generalized estimating equation approach to examine time-to-event outcomes including symptom duration, shedding duration, and time to infection. The duration of symptoms varied by bacterial community type both prior to and during influenza virus infection. Further, a community type with low diversity was associated with shorter duration of viral shedding and delayed time to infection among secondary cases. The results of these various analyses suggest the respiratory microbiome may be a potential target for reducing influenza risk, household transmission, and disease severity. In the final chapter, I review the skills I learned and the challenges I encountered during the dissertation process. Finally, I review future research directions that focus on deciphering the complex dynamics between the host, pathogen, and microbiome.